Skip to main content

Table 2 Stepwise multivariate logistic regression for occurrence of a major surgical resection using RHA group

From: Does geography influence the treatment and outcomes of colorectal cancer? A population-based analysis

Variable Adjusted ORa 95% CI P value
Site    <0.0001
  Colon 1   
  Rectosigmoid 0.76 0.51 to 1.14  
  Rectum 0.53 0.40 to 0.70  
AJCC stage    <0.0001
  I 1   
  II 4.11 2.83 to 5.97  
  III 5.97 3.87 to 9.20  
  IV 0.34 0.24 to 0.48  
  Unknown/NAb 0.09 0.04 to 0.19  
Income quintilec    0.045
  Quintile1 1   
  Quintile 2 1.04 0.73 to 1.48  
  Quintile 3 1.37 0.94 to 1.98  
  Quintile 4 1.41 0.96 to 2.07  
  Quintile 5 1.26 0.85 to 1.86  
  NF 0.32 0.11 to 0.93  
CCI group    0.045
  CCI count <1 1   
  CCI count = 1 1.39 1.00 to 1.93  
  CCI count >1 0.96 0.68 to 1.35  
RHA group    0.037
  Winnipeg 1   
  North Manitoba 0.48 0.26 to 0.90  
  South Manitoba 1.25 0.93 to 1.69  
  Middle Manitoba 1.11 0.78 to 1.57  
  1. Potential covariates in the model included: age group (<70 years, >70 years), gender, site (colon, rectosigmoid, rectum), AJCC stage, income quintile, Charlson comorbidity index (CCI) group, and RHA group. Age and gender were not significant in the stepwise analysis.
  2. aThe higher the odds ratio, the higher the odds of undergoing major surgery; bone or more elements of the AJCC stage is missing or staging scheme is NA; cquintile 1: poorest, quintile 5: richest, NF: patients for whom income quintile information was not found, including patients living in institutionalized facilities, such as personal care homes, mental health institutes, prisons, or offices of the public trustee. AJCC American Joint Committee on Cancer, CCI Charlson comorbidity index, CI confidence interval, NA not applicable, NF not formatted, OR odds ratio, RHA regional health authority.