ARTICLE SUPPLEMENTARY INFORMATION
(Supporting information is provided by the author to support the content
of the parent article)

The Impact of Drug Vintage on Patient Survival: A Patient-Level Analysis Using Quebec's Provincial Health Plan Data

Frank R. Lichtenberg, Paul Grootendorst, Marc Van Audenrode, Dominick Latremouille-Viau, Patrick Lefebvre

– Value in Health Supplemental Information

  1. Logistic regression modeling the probability of dying

Variable

Odds Ratio

95 %

Confidence Limit

P-Value

Drug vintage

 

 

 

 Pre-1970 (reference)

-

-

-

 Post-1970

1.228

[1.110 ; 1.358]

<.0001

 Post-1980

0.735

[0.657 ; 0.822]

<.0001

 Post-1990

0.511

[0.461 ; 0.566]

<.0001

Demographics

 

 

 

Age

1.222

[1.208 ; 1.237]

<.0001

Female

0.581

[0.558 ; 0.604]

<.0001

Year dummies

 

 

 

1997 (reference)

-

-

-

1998

1.032

[0.901 ; 1.182]

0.6521

1999

1.675

[1.483 ; 1.895]

<.0001

2000

1.605

[1.419 ; 1.817]

<.0001

2001

2.038

[1.805 ; 2.305]

<.0001

2002

1.982

[1.748 ; 2.249]

<.0001

2003

1.787

[1.568 ; 2.039]

<.0001

2004

1.691

[1.475 ; 1.941]

<.0001

2005

1.428

[1.237 ; 1.651]

<.0001

2006

1.118

[0.961 ; 1.302]

0.1514

Government GIS status

 

 

 

Full GIS

2.228

[2.039 ; 2.431]

<.0001

Partial GIS

1.010

[0.969 ; 1.053]

0.6263

Without GIS (reference)

-

-

-

Region

 

 

 

Montreal

0.905

[0.861 ; 0.950]

<.0001

Quebec

1.044

[0.973 ; 1.121]

0.2293

Mauricie

1.019

[0.946 ; 1.096]

0.6227

Monteregie

1.012

[0.956 ; 1.070]

0.6834

Other (reference)

-

-

-

Yearly medical resources utilization

 

 

 

Inpatient hospitalization frequency

1.745

[1.716 ; 1.773]

<.0001

Inpatient length of stay

1.027

[1.026 ; 1.028]

<.0001

Outpatient consultation frequency

0.984

[0.983 ; 0.986]

<.0001

Specialist consultation (yes/no)

1.077

[1.001 ; 1.159]

0.0471

Number of prescription

0.976

[0.976 ; 0.977]

<.0001

Comorbidities

 

 

 

Number of disease category

 

 

 

    No disease (reference)

-

-

-

    1 disease

1.736

[1.283 ; 2.326]

0.0003

    2 diseases

2.253

[1.666 ; 3.016]

<.0001

    3 diseases

2.849

[2.106 ; 3.815]

<.0001

    4 diseases

3.453

[2.550 ; 4.633]

<.0001

    ≥5 diseases

4.510

[3.326 ; 6.059]

<.0001

Disease category

 

 

 

    Infectious and parasitic diseases

0.045

[0.031 ; 0.066]

<.0001

    Neoplasms

8.545

[6.488 ; 11.385]

<.0001

    Endocrine, nutritional, metabolic, immunity

0.266

[0.197 ; 0.363]

<.0001

    Mental disorders

0.325

[0.238 ; 0.447]

<.0001

    Nervous system and sense organs

0.105

[0.077 ; 0.145]

<.0001

    Circulatory system

0.649

[0.492 ; 0.865]

0.0027

    Respiratory system

0.898

[0.673 ; 1.210]

0.4712

    Digestive system

0.213

[0.153 ; 0.299]

<.0001

    Genitourinary system

5.960

[4.382 ; 8.139]

<.0001

    Musculoskeletal system and connective tissue

1.138

[0.825 ; 1.573]

0.4330

    Symptoms, signs, and ill-defined conditions

34.162

[25.943 ; 45.268]

<.0001

2. Accelerated failure time model using Weibull distribution

Variable

Parameter Estimate

95 %

Confidence Limit

P-Value

Drug vintage

 

 

 

 Pre-1970 (reference)

-

-

-

 Post-1970

-0.012

[-0.031 ; 0.008]

0.2365

 Post-1980

0.027

[0.005 ; 0.048]

0.0138

 Post-1990

0.022

[0.003 ; 0.041]

0.0268

Demographics

 

 

 

Age

0.015

[0.013 ; 0.016]

<.0001

Female

0.042

[0.036 ; 0.048]

<.0001

Year dummies

 

 

 

1997 (reference)

-

-

-

1998

0.753

[0.732 ; 0.774]

<.0001

1999

1.191

[1.173 ; 1.210]

<.0001

2000

1.488

[1.469 ; 1.507]

<.0001

2001

1.715

[1.697 ; 1.734]

<.0001

2002

1.892

[1.874 ; 1.911]

<.0001

2003

2.039

[2.020 ; 2.059]

<.0001

2004

2.163

[2.143 ; 2.183]

<.0001

2005

2.280

[2.259 ; 2.301]

<.0001

2006

2.965

[2.940 ; 2.989]

<.0001

Government GIS status

 

 

 

Full GIS

0.052

[0.039 ; 0.065]

<.0001

Partial GIS

0.050

[0.044 ; 0.056]

<.0001

Without GIS (reference)

-

-

-

Region

 

 

 

Montreal

0.005

[-0.003 ; 0.012]

0.8855

Quebec

-0.001

[-0.011 ; 0.010]

0.6574

Mauricie

-0.003

[-0.014 ; 0.009]

0.4306

Monteregie

-0.003

[-0.012 ; 0.005]

<.0001

Other (reference)

-

-

-

Yearly medical resources utilization

 

 

 

Inpatient hospitalization frequency

-0.013

[-0.015 ; -0.011]

<.0001

Inpatient length of stay

0.000

[0.000 ; 0.000]

0.5718

Outpatient consultation frequency

0.001

[0.001 ; 0.001]

<.0001

Specialist consultation (yes/no)

0.005

[-0.007 ; 0.017]

0.4095

Number of prescription

0.001

[0.001 ; 0.002]

<.0001

Comorbidities

 

 

 

Number of disease category

 

 

 

    No disease (reference)

-

-

-

    1 disease

-0.051

[-0.101 ; 0.000]

0.0508

    2 diseases

-0.052

[-0.102 ; -0.001]

0.0446

    3 diseases

-0.068

[-0.118 ; -0.017]

0.0086

    4 diseases

-0.080

[-0.131 ; -0.030]

0.0018

    ≥5 diseases

-0.088

[-0.138 ; -0.038]

0.0006

Disease category

 

 

 

    Infectious and parasitic diseases

0.150

[0.091 ; 0.209]

<.0001

    Neoplasms

-0.037

[-0.085 ; 0.011]

0.1272

    Endocrine, nutritional, metabolic, immunity

0.090

[0.039 ; 0.141]

0.0006

    Mental disorders

0.107

[0.055 ; 0.158]

<.0001

    Nervous system and sense organs

0.180

[0.127 ; 0.232]

<.0001

    Circulatory system

0.023

[-0.025 ; 0.071]

0.3493

    Respiratory system

-0.008

[-0.057 ; 0.042]

0.7559

    Digestive system

0.066

[0.012 ; 0.121]

0.0173

    Genitourinary system

-0.130

[-0.174 ; -0.086]

<.0001

    Musculoskeletal system and connective tissue

0.031

[-0.015 ; 0.077]

0.1819

    Symptoms, signs, and ill-defined conditions

-0.148

[-0.187 ; -0.108]

<.0001

 

Note that, as opposed to the Cox model where a positive estimate denotes an increased probability of dying (i.e., decreased survival duration), in accelerated failure time model, a negative estimate is indicative of decreased survival duration.  For example, the positive coefficient for the variable “female” indicates that women had longer survival duration than men.  As in any regression analysis, these coefficients adjust or control for all other covariates in the model.

3. Accelerated failure time model using Gamma distribution

Variable

Parameter Estimate

95 %

Confidence Limit

P-Value

Drug vintage

 

 

 

 Pre-1970 (reference)

-

-

-

 Post-1970

-0.002

[-0.008 ; 0.005]

0.6466

 Post-1980

0.010

[0.002 ; 0.017]

0.011

 Post-1990

0.011

[0.003 ; 0.018]

0.0055

Demographics

 

 

 

Age

0.002

[0.001 ; 0.003]

<.0001

Female

0.007

[0.004 ; 0.009]

<.0001

Year dummies

 

 

 

1997 (reference)

-

-

-

1998

0.692

[0.685 ; 0.699]

<.0001

1999

1.096

[1.089 ; 1.102]

<.0001

2000

1.381

[1.375 ; 1.388]

<.0001

2001

1.607

[1.600 ; 1.613]

<.0001

2002

1.785

[1.778 ; 1.792]

<.0001

2003

1.934

[1.927 ; 1.941]

<.0001

2004

2.068

[2.061 ; 2.075]

<.0001

2005

2.186

[2.178 ; 2.194]

<.0001

2006

3.080

[3.060 ; 3.100]

<.0001

Government GIS status

 

 

 

Full GIS

0.014

[0.009 ; 0.020]

<.0001

Partial GIS

0.012

[0.010 ; 0.015]

<.0001

Without GIS (reference)

-

-

-

Region

 

 

 

Montreal

0.004

[0.001 ; 0.007]

0.0156

Quebec

-0.001

[-0.005 ; 0.003]

0.6788

Mauricie

-0.003

[-0.007 ; 0.001]

0.15

Monteregie

0.002

[-0.001 ; 0.006]

0.1436

Other (reference)

-

-

-

Yearly medical resources utilization

 

 

 

Inpatient hospitalization frequency

-0.003

[-0.003 ; -0.002]

<.0001

Inpatient length of stay

0.000

[0.000 ; 0.000]

0.5126

Outpatient consultation frequency

0.000

[0.000 ; 0.001]

<.0001

Specialist consultation (yes/no)

0.005

[-0.001 ; 0.010]

0.0874

Number of prescription

0.001

[0.000 ; 0.001]

<.0001

Comorbidities

 

 

 

Number of disease category

 

 

 

    No disease (reference)

-

-

-

    1 disease

-0.031

[-0.050 ; -0.011]

0.0019

    2 diseases

-0.032

[-0.051 ; -0.013]

0.0013

    3 diseases

-0.036

[-0.056 ; -0.017]

0.0003

    4 diseases

-0.041

[-0.061 ; -0.021]

<.0001

    ≥5 diseases

-0.043

[-0.062 ; -0.023]

<.0001

Disease category

 

 

 

    Infectious and parasitic diseases

0.056

[0.031 ; 0.081]

<.0001

    Neoplasms

-0.006

[-0.023 ; 0.011]

0.4787

    Endocrine, nutritional, metabolic, immunity

0.036

[0.016 ; 0.056]

0.0004

    Mental disorders

0.068

[0.047 ; 0.089]

<.0001

    Nervous system and sense organs

0.087

[0.064 ; 0.110]

<.0001

    Circulatory system

0.013

[-0.003 ; 0.030]

0.1173

    Respiratory system

-0.002

[-0.019 ; 0.015]

0.8067

    Digestive system

0.033

[0.013 ; 0.054]

0.0013

    Genitourinary system

-0.040

[-0.062 ; -0.018]

0.0004

    Musculoskeletal system and connective tissue

0.022

[-0.003 ; 0.047]

0.0882

    Symptoms, signs, and ill-defined conditions

-0.042

[-0.062 ; -0.023]

<.0001

 

Note that, as opposed to the Cox model where a positive estimate denotes an increased probability of dying (i.e., decreased survival duration), in accelerated failure time model, a negative estimate is indicative of decreased survival duration.  For example, the positive coefficient for the variable “female” indicates that women had longer survival duration than men.  As in any regression analysis, these coefficients adjust or control for all other covariates in the model.

4. Accelerated failure time model using log-logistic distribution

Variable

Parameter Estimate

95 %

Confidence Limit

P-Value

Drug vintage

 

 

 

 Pre-1970 (reference)

-

-

-

 Post-1970

-0.021

[-0.046 ; 1.005]

0.1124

 Post-1980

0.040

[0.011 ; 0.068]

0.0069

 Post-1990

0.026

[0.002 ; 0.051]

0.0378

Demographics

 

 

 

Age

0.027

[0.025 ; 0.029]

<.0001

Female

0.068

[0.060 ; 0.076]

<.0001

Year dummies

 

 

 

1997 (reference)

-

-

-

1998

0.927

[0.891 ; 0.964]

<.0001

1999

1.377

[1.344 ; 1.411]

<.0001

2000

1.683

[1.650 ; 1.717]

<.0001

2001

1.899

[1.866 ; 1.932]

<.0001

2002

2.069

[2.036 ; 2.103]

<.0001

2003

2.209

[2.175 ; 2.243]

<.0001

2004

2.321

[2.286 ; 2.356]

<.0001

2005

2.430

[2.394 ; 2.466]

<.0001

2006

3.043

[3.007 ; 3.080]

<.0001

Government GIS status

 

 

 

Full GIS

0.073

[0.056 ; 0.090]

<.0001

Partial GIS

0.078

[0.071 ; 0.086]

<.0001

Without GIS (reference)

-

-

-

Region

 

 

 

Montreal

0.001

[-0.008 ; 0.011]

0.7793

Quebec

0.002

[-0.012 ; 0.015]

0.8072

Mauricie

-0.002

[-0.016 ; 0.012]

0.7772

Monteregie

-0.007

[-0.018 ; 0.004]

0.2205

Other (reference)

-

-

-

Yearly medical resources utilization

 

 

 

Inpatient hospitalization frequency

-0.026

[-0.029 ; -0.022]

<.0001

Inpatient length of stay

0.000

[0.000 ; 0.000]

0.4204

Outpatient consultation frequency

0.001

[0.001 ; 0.001]

<.0001

Specialist consultation (yes/no)

0.008

[-0.006 ; 0.022]

0.2379

Number of prescription

0.002

[0.002 ; 0.002]

<.0001

Comorbidities

 

 

 

Number of disease category

 

 

 

    No disease (reference)

-

-

-

    1 disease

-0.038

[-0.100 ; 0.024]

0.2298

    2 diseases

-0.038

[-0.100 ; 0.024]

0.2314

    3 diseases

-0.056

[-0.118 ; 0.006]

0.0761

    4 diseases

-0.071

[-0.133 ; -0.009]

0.0255

    ≥5 diseases

-0.082

[-0.144 ; -0.020]

0.0093

Disease category

 

 

 

    Infectious and parasitic diseases

0.177

[0.105 ; 0.249]

<.0001

    Neoplasms

-0.077

[-0.136 ; -0.017]

0.0114

    Endocrine, nutritional, metabolic, immunity

0.095

[0.032 ; 0.158]

0.003

    Mental disorders

0.090

[0.026 ; 0.153]

0.0058

    Nervous system and sense organs

0.180

[0.117 ; 0.244]

<.0001

    Circulatory system

0.023

[-0.037 ; 0.082]

0.4528

    Respiratory system

-0.011

[-0.073 ; 0.050]

0.7181

    Digestive system

0.068

[0.000 ; 0.136]

0.0491

    Genitourinary system

-0.142

[-0.194 ; -0.090]

<.0001

    Musculoskeletal system and connective tissue

0.014

[-0.039 ; 0.067]

0.6039

    Symptoms, signs, and ill-defined conditions

-0.192

[-0.241 ; -0.144]

<.0001

 

Note that, as opposed to the Cox model where a positive estimate denotes an increased probability of dying (i.e., decreased survival duration), in accelerated failure time model, a negative estimate is indicative of decreased survival duration.  For example, the positive coefficient for the variable “female” indicates that women had longer survival duration than men.  As in any regression analysis, these coefficients adjust or control for all other covariates in the model.

 

Value in Health Article Supplementary Information
 

Contact ISPOR @ info@ispor.org  |  View Legal Disclaimer
©2010 International Society for Pharmacoeconomics and Outcomes Research. All rights reserved under International and Pan-American Copyright Conventions. 
Website design by Eagle Systems USA, Inc.