Skip to main content

Table 4 Simple Poisson Multivariable Regression Model of Steroid Exposure

From: Examining systemic steroid use in older inflammatory bowel disease patients using hurdle models: a cohort study

  Poisson regression
  Duration of Steroid Use
  RR CI
Age ( 5 year decrease) 1.19 (1.08, 1.31)e
Female 1.40 (0.97, 2.02)
Caucasian 1.39 (0.80, 2.42)
Region   
 Midwest 1.00 Reference
 Northeast 0.86 (0.56, 1.32)
 Southeast 0.71 (0.46, 1.08)
 Southwest 0.60 (0.30, 1.18)
 Rocky Mountain 0.42 (0.09, 2.05)
 Pacific & HI, AK, Prf 0.60 (0.29, 1.23)
Rural 1.23 (0.86, 1.77)
Medicaid eligible 1.04 (0.72, 1.51)
antiTNFs 0.62 (0.19, 2.03)
Nonbiologic immunomodulators 1.58 (0.75, 3.33)
Aminosalicylates 1.41 (0.98, 2.03)
Locally administered steroids 4.39 (1.51, 12.70)d
Antidiarrheals 1.07 (0.69, 1.65)
Charlson index 0.96 (0.90, 1.01)
Hospitalizations 0.96 (0.87, 1.05)
Endoscopy (>1) 0.87 (0.60, 1.27)
IBD-associated Surgery 0.68 (0.35, 1.33)
  1. aDependent variable was days of steroid use among participants (n = 1216) during the follow-up period; The models were run with robust standard errors, Individuals with baseline steroid use were excluded from these models; bRR = rate ratios; All explanatory variables are measured within the first 12 months after study entry; c p < .05, d p < .01, e p < .001; fHI = Hawaii, AK = Alaska, PR = Puerto Rico