Transcript Media exposure - The Conference Exchange

THE EFFECTS OF MEDIA EXPOSURE IN DIRECTTO CONSUMER ADVERTISING OF
PRESCRIPTION DRUGS, PATIENT DEMAND,
AND PATIENT SATISFACTION
140th Annual American Public Health Association (APHA)
Conference
San Francisco, California
October 31,2012
Marvin Rock, Mian B. Hossain1, Andrea Kidd-Taylor1, H. Eduardo Velasco2
1 School
2
of Community of Health and Policy, Morgan State University
College of Osteopathic Medicine, Touro University
Presentation Outline
2








Rationale of the study
Direct-to Consumer Advertising Definition
Research Question & Hypotheses
Conceptual Model & Adapted Model
Data Source
Univariate, Bivariate, Multivariate Models
Analytical Model
Conclusions
Rationale for the Study
3

Why Prescription Drugs?
 Cost (fastest rising health care expense)
 Leading to an increase in new medicines being
approved and marketed
 Consumer Demand is growing for new medications
 Intensification of prescription drug marketing has lead
to DTCA
Congressional Budget Office, 2006
Direct to Consumer Advertising (DTCA)-Defined

DTCA : the promotion of prescription drugs directly to the consumer
through the following methods:

Sources of DTCA:

Newspaper

Magazine

Television or radio

Internet marketing (Social Marketing i.e. Facebook, Twitter, and
Blogs )
Print
Broadcast
Online

Family & Friends

Healthcare providers
Other
Consumer for Media and Democracy, 2007
Research Question and Research Hypotheses
5

RQ1: What are the health determinants of patient
demand?

Relationship between differing levels of media exposure of
DTCA prescription drugs
H1: Do differing levels of media exposure (hi, med, lo)
affect patient demand
 H2: Does health insurance status moderate the effects of
media exposure and patient demand
 H3: Are patients with higher levels of media exposure
more likely to have patient satisfaction adjusting for all
other variables

Bero’s Conceptual Model: (2001)
2. Consumer Attitudes:
•Consumer Assess. of DTC Ads
•Profit Motivated Ads vs. Public
Service Announcements
•Promotion vs. Education
•Medicalization of Everyday
Problems
1. DTC Advertising
• Exposure (Ad
Expenditures)
• Content
• Accuracy
9. Health Care Costs
• Drug Costs to Consumers &
Health Plans
• Physician Visits
• Diagnostic Tests
• Hospitalization
3. Patient Demand
5. Patient-Physician
Relationship
• Satisfaction
• Trust
• Visit Efficiency
6. Prescribing
(Physician Actions)
• Quality of Prescribing
• Quantity of Generic Drug Prescribe
• Quantity of Brand Name Drug
Prescribed
(Physician Group/HMO Actions)
• Formulary Compliance
• Economic Viability of the Group /HMO
7. Adherence
with Drug
Therapy
8.Patients’
Health
Outcomes
• Perceived
Health
Status
• Objective
Measures
of Health
Status
10. Pharmaceutical Industry
• Market Share
• Profit Level
4. Physician Attitudes
Bero, 2001
Adapted Conceptual Model: (Rock)
Patient Demand
DTC Advertising
• Media Exposure
Patients’ Health Outcomes
• Adverse Drug Events
• Patient Satisfaction
• Health Related QoL
Rock, 2011
“Public Health Impact of Direct to Consumer
Advertising of Prescription Drugs”(ICPSR 3687)
8



Cross-sectional study design
Secondary data taken from 2001/2002 Public Health
Impact of Direct-to-Consumer Advertising of Prescription
Drugs (n=3000)
Twenty minute telephone interview, with a nationally
representative sample (telephone households in
Continental United States)



Sample size (3000), 18 years of age and older
Stratified sampling process
Households selected computerized by Random Digit Dialing
(RDD)




Genesys Sampling System
Based on # households in the exchange
Proper representations of households in different regions
Central city, suburban, rural
Weissman et al., 2003
Study Covariates
9
Variables
Type
Health status
Categorical
Gender
Dichotomous
Age
Categorical
Education
Categorical
Health insurance status
Dichotomous
Marital status
Categorical
Income
Categorical
Employment status
Dichotomous
Geographic region
Categorical
Health Related Quality of Life
Dichotomous
Race/ethnicity
Categorical
Adverse Drug Events
Dichotomous
Weighted socio-demographic and Health characteristics of
the study population
Variable
Proportion
Sample size
(n)
Socio-demographic variables
Media exposure (n=2,957)
Low media exposure
0.35
992
Medium media exposure
0.34
989
High media exposure
0.47
976
Television or radio
0.15
412
Internet
0.13
359
Mass media & print media
0.14
412
Print media
0.17
522
Family & Friends
0.16
480
Health provider
0.24
694
Overall Exposure to health information (n=2,879)
10
Weighted distributions of patient demand and patient satisfaction,
two outcome variables for the study
Variable
Proportion
Sample size (n)
Yes
0.31
801
No
0.69
1,791
Better
0.81
520
Worse
0.19
130
Patient Demand
Advertisement prompted to talk to Dr. (n=2,592)
Patient Satisfaction
Overall do you feel better after taking prescribed
drug (n=650)
11
Bivariate Association Between Media Exposure and
Other Covariates with Patient Demand
Independent Variables
Media Exposure
Patient Demand
S
Geographic Region
NS
Income
NS
Health insurance
NS
Gender
S
Race/Ethnicity
NS
Marital Status
NS
Employment Status
NS
Age
NS
Education
NS
Health Status
S
Geographic Location
S
S: Statistically
significant with at least
95% confidence.
NS: Not statistically
significant.
Bivariate Association Between Media Exposure and
Other Covariates with Patient Satisfaction
Independent Variables
Patient Satisfaction
Media Exposure
NS
Geographic Region
NS
Income
S
Health Insurance
NS
Gender
NS
Race/Ethnicity
S
Marital Status
NS
Employment Status
NS
Age
NS
Education
NS
Health Status
NS
Geographic location
NS
HRQL
S
ADE
S
S: Statistically
significant with at least
95% confidence.
NS: Not statistically
significant.
Analytical Models
14


Logistic regression- Patient Demand (yes/no) and Patient Satisfaction
(better/worse)
Model 1-3:








Media exposure (TV, radio, print, and Internet)
Patient demand
Socio-demographic variables
Health status
Model 4: Model 1+interaction between health insurance status
Model 5-7: Media exposure & patient satisfaction controlling for SD, HS,
Adverse Drug Events, and Health Relate Quality of Life
Model 8: Model 5+interaction between health insurance status
Odds ratios


Less likely to demand & less satisfied OR<1 PD(no)/ PS (worse)
More likely to demand & more satisfied OR>1 PD(yes)/ PS (better)
Logistic regression models estimates for the relationship between
patient’s levels of media exposure and patient demand
Covariates
Model 1
Model 2
Model 3
OR
95% CI
OR
95% CI
OR
95% CI
Low media exposure (RC)
1.00
-
1.00
-
1.00
-
Medium media exposure
1.53***
1.11, 3.48 1.54***
20,1.97
1.51**
1.16, 1.97
High media exposure
2.32***
95,2.78
2.33***
1.83, 2.96 2.30***
1.77, 2.98
Insured (RC)
-
-
1.00
-
-
Uninsured
-
-
0.86
0.63, 1.18 0.79
Media exposure
Health Insurance
Significance: * p<0.05; ** p<0.01; *** p<0.001
15
1.00
0.56, 1.13
Logistic regression models estimates for the relationship between patient’s
levels of media exposure, patient demand ,
and the interaction term
Covariates
Model 4
OR
95% CI
Low media exposure (RC)
1.00
-
Medium Media exposure
2.14
0.93. 4.90
2.71**
1.22, 5.99
Insured (RC)
1.00
-
Uninsured
0.64
0.33, 1.26
-
-
1.00
-
Media exposure
High media exposure
Health Insurance
Media exposure x Health Insurance status
int_media_exp_low_new_healthins2
16
int_media_exp_low_new_healthins1(RC)
Logistic regression estimates for the relationship between media
exposure and patient’s satisfaction controlling for all other variables
Covariates
Model 5
Model 6
Model 7
OR
95% CI
OR
95% CI
OR
95% CI
Low media exposure (RC)
1.00
-
1.00
-
1.00
-
Medium media exposure
1.96*
1.11, 3.48
1.98**
1.12, 3.50
2.14*
1.08, 4.22
High media exposure
1.63
0.95, 2.78 1.63
0.95,2.79
1.83
0.94,3.60
Yes (RC)
-
-
-
-
1.00
-
No
-
-
-
-
3.25***
1.90, 5.56
Decreased ability (RC)
-
-
-
-
1.00
-
Increased ability
-
-
-
-
5.80***
2.69,12.50
Intercept
1.03
0.967
0.077
Sample Size
643
643
583
Media exposure
Experiencing any side effects from taking rx
drug
Prescribed drug had effect on your ability
17
Significance: * p<0.05; ** p<0.01; *** p<0.001
Conclusions



Patients with higher levels of exposure to media
(medium and high exposure) are significantly more likely
to have patient demand
The interaction between patient demand and health
insurance status was not a significant factor or
moderator in the media exposure patient demand
relationship
Patients with medium level of media exposure are
significantly more likely to have patient satisfaction,
adjusting for adverse drug events and health related
quality of life, socio-demographic variables, and health
status
Future Implications
19


Further need to conduct research on DTCA of
prescription drugs & patient health outcomes
Readdressing current policies on this form of
health communication
Risk/benefit drug information available to patients
 Post approval risk/benefit analysis
 Quicker evaluation of drug ads that violate quality of life
and economic claims
 Reevaluation of the current regulations on DTCA

Acknowledgements
20
Thank you








Dr. Mian Hossain
Dr. Andrea Kidd-Taylor
Dr. Neil Alperstein
Dr. Lisa Bero
Dr. Mary Carter
Dr. Barbara Mintzes
Dr. Joel Weissman
Dr. Lester Spence
Thank You





Dr. Eduardo Velasco
Dr. Timothy Akers
Dr. Jonathan VanGeest
Dr. Robert Jagers
University of Michigan

Inter-university
Consortium for Political
and Social Research
21
Contact Information
Marvin A. Rock, Dr.P.H., M.P.H.
Health Outcomes Scientist/ Epidemiologist
[email protected]
Questions?
22