Skip to main content

Table 3 Estimates of the generalized linear mixed models on the probability of generic and branded substitution

From: Analyzing generic and branded substitution patterns in the Netherlands using prescription data

 

Generic substitution

Branded substitution

 

Log Odds (β)

P-value

Log Odds (β)

P-value

Fixed effects estimates

    

Intercept

4.262

< 0.001

-6.978

< 0.001

Prescriptions to female patients

-0.163

0.284

0.095

0.391

First prescription in the drug class

-2.582

< 0.001

2.752

< 0.001

Prescription history

0.076

< 0.001

0.016

0.361

Branded ratio

    

Branded ratio [0 - 0.33)

baseline

Branded ratio [0.33 - 0.66)

-3.210

< 0.001

2.960

< 0.001

Branded ratio [0.66 - 1]

-5.952

< 0.001

5.809

< 0.001

Pharmacy property status

    

Individually owned pharmacy

baseline

Pharmacy in a partnership

0.954

0.046

-1.752

< 0.001

Pharmacy in a chain

0.687

< 0.001

-0.605

< 0.001

Duration after patent expiry (months)

-0.011

0.002

-0.046

< 0.001

Age

    

Age [16-31]

baseline

Age (31-46]

0.568

< 0.007

-0.399

0.038

Age (46-52]

0.546

0.012

-0.146

0.307

Age (52-88]

0.474

0.028

-0.153

0.262

Drug group

    

   ACE Inhibitors

baseline

   Antidepressants

-0.579

0.004

1.846

< 0.001

   PPIs

-1.085

< 0.001

0.986

< 0.001

   Statins

0.574

0.001

-0.634

< 0.001

Random effects estimates

    

Pharmacy-level residual variance

1.915

0.350

Patient- level residual variance

2.346

1.879

Somers D xy criterion

0.985

0.981