Skip to main content

Table 1 Multivariate logistic regression of factors associated with severe liver disease

From: Practical considerations for sensitivity analysis after multiple imputation applied to epidemiological studies with incomplete data

    

Multivariate analysis

Factors

Patients (n = 4 343)

% SLD

% missing data

Complete Case (n = 2 130) aOR* (95% CI*)

Multiple Imputation (n = 4 343) aOR* (95% CI*) M = 30 imputed datasets

Period of inclusion

     

2001-2003

2330

7.0

   

2004-2007

2013

9.5

   

Sex

     

Female

993

4.2

 

1.0

1.0

Male

3350

9.3

 

1.8 [1.1,3.0]

2.0 [1.4,2.9]

Age

     

≤ 40 years

2435

3.9

 

1.0

1.0

> 40 years

1908

13.6

 

2.2 [1.5,3.3]

2.3 [1.7,3.1]

Time between 1st HCV + test and referral

     

< 1 year

1728

6.7

   

≥ 1 year

2163

8.7

   

Missing

452

11.5

10.4

  

Duration of HCV infection at referral

     

< 18 years

1709

3.0

 

1.0

1.0

≥ 18 years

2002

12.5

 

3.1 [2.0,5.1]

2.6 [1.8,3.7]

Missing

632

8.2

14.6

  

History of excessive alcohol intake

     

No

2015

4.5

 

1.0

1.0

Yes

1847

13.2

 

2.6 [1.8,3.7]

2.8 [2.2,3.7]

Missing

481

4.4

11.1

  

HbsAg status

     

Negative

3570

8.3

 

1.0

 

Positive

89

13.5

 

2.4 [1.0,5.9]

 

Missing

684

6.7

15.7

  

HIV serostatus

     

Negative

3342

8.2

  

1.0

Positive

294

14.0

  

1.8 [1.2,2.6]

Missing

707

5.7

16.3

  

HCV genotype 3

     

No

2083

7.2

 

1.0

1.0

Yes

1117

10.3

 

1.5 [1.1,2.0]

1.6 [1.3,2.1]

Missing

1143

7.8

26.3

  
  1. *aOR, adjusted Odds Ratio; CI, confidence interval; SLD, severe liver disease (cirrhosis, hepatocellular carcinoma); HCV, hepatitis C virus; HBsAg, hepatitis B surface antigen; HIV, human immunodeficiency virus.
  2. † Time from suspected year of infection to year of referral to the reference centre. Suspected year of HCV infection is defined as year of the last HCV negative test performed during the drug-use period or year of first drug injection.
  3. ‡ >210 g/week for women and >280 g/week for men.
  4. Complete case and multiple imputation analyses were applied to a population of HCV-RNA positive drug users newly referred in hepatology reference centres in France, 2001–2007.