Skip to main content

Table 2 Regressions among the matched sample (N = 2,116)—Model M1

From: Assessing real-world effectiveness of therapies: what is the impact of incretin-based treatments on hospital use for patients with type 2 diabetes?

  

Probability of being hospitalized

Length of stay

  

Linear probability model

Zero truncated Poisson

Variables

 

Coeff

 

95%CI

IRR

 

95%CI

DID

 

-0,01

 

-0,04

-

0,03

0,62

***

0,57

-

0,68

Post period

 

0,04

 

0,00

-

0,07

1,24

***

1,24

-

1,35

Treatment (Incretin-based treatment)

1

    

1

    

hospital use in the previous year

 

0,13

***

0,11

-

0,16

1,03

 

0,98

-

1,07

At least one endocrinologist visit

0,05

**

0,00

-

0,10

1,03

 

0,93

-

1,15

Type of therapy for diabetes

          
 

Monotherapy

Ref

    

Ref

    
 

Dual therapy

0,01

 

-0,02

-

0,04

1,35

***

1,25

-

1,46

 

Triple therapy

0,03

 

-0,01

-

0,07

1,61

***

1,46

-

1,77

 

Insulin therapy

0,19

***

0,13

-

0,26

1,98

***

1,80

-

2,18

Diabetes follow-up (score)

-0,05

**

-0,09

-

-0,02

0,83

**

0,74

-

0,93

Polypharmacy

0,04

***

0,04

-

0,05

1,03

***

1,02

-

1,03

Age

 

0,00

**

0,00

-

0,00

1,04

**

1,01

-

1,07

Living with partner

0,00

 

-0,02

-

0,01

1,05

 

0,55

-

2,03

Working

 

-0,03

 

-0,03

-

0,05

0,77

*

0,58

-

1,01

100% coverage status

 

0,04

**

-0,11

-

0,03

1,66

***

1,37

-

2,00

Ln(income)

 

-0,04

 

-0,01

-

0,02

1,42

**

1,08

-

1,86

Location in a rural area

-0,01

 

-0,24

-

0,23

0,50

**

0,27

-

0,94

Share of diabetes-related deaths

-0,02

 

-0,05

-

0,02

1,55

**

1,33

-

1,81

Density of hospital beds

0,00

 

0,00

-

0,00

1,01

***

1,01

-

1,01

  1. Notes. Coeff = coefficient of the linear probability model; IRR = incidence rate ratios; 95%CI = 95% confidence intervals; * p < 0.1, ** p < 0.05, *** p < 0.0001
  2. Second part of the Hurdle model, i.e. the number of days hospitalized conditionally to at least one hospital admission
  3. Effect of treatment omitted because of no within group variation