Skip to main content

Table 2 Predicted probabilities of CRC incidence and mortality in the two cohorts - Multivariate logistic regression model

From: Methodological approaches to population based research of screening procedures in the presence of selection bias and exposure measurement error: colonoscopy and colorectal cancer outcomes in Ontario

Study cohorts

 

Colonoscopy

Estimate

95% CI

Absolute risk difference

95% CI

Unselected cohort (n = 1,341,612)

      
 

Incidence

     
  

No

1.75

1.73, 1.77

3.26

3.22, 3.30

  

Yes

5.01

4.97, 5.04

  
 

Mortality

     
  

No

0.38

0.33, 0.42

0.79

0.71, 0.87

  

Yes

1.17

1.10, 1.23

  

Restricted cohort (n = 1,089,998)

      
 

Incidence

     
  

No

1.42

1.39, 1.44

-0.53

-0.60, -0.44

  

Yes

0.90

0.82, 0.98

  
 

Mortality

     
  

No

0.24

0.18, 0.29

-0.08

-0.19, 0.04

  

Yes

0.16

0.13, 0.34

  
  1. Data are presented as percentages.
  2. CRC, colorectal cancer; CI, confidence interval.
  3. Multivariate logistic model adjusted subject characteristics such as age, sex, charlson comorbidity index, rural area of residence and income quintile.
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us