Trends in the Pharmaceutical Treatment of Diabetes and Implications for Financial Burdens, 1997-2007 (Text Version)

On September 27, 2010, Eric Sarpong, made this presentation at the 2010 Annual Conference. Select to access the PowerPoint® presentation (1.79 MB).  

---

Slide 1

Trends in the Pharmaceutical Treatment of Diabetes and Implications for Financial Burdens, 1997-2007

Eric Sarpong, Didem M. Bernard and G. Edward Miller
AHRQ Annual Conference.
September 27, 2010
 

Slide 2

Overview

• Diabetes is a chronic, progressive and costly disease (2007, $174 billion)
• Trends in anti-diabetic drugs are important
  • Increasing disease prevalence
  • Chronic comorbidities
  • More aggressive treatments
  • Newer and more expensive drugs
• Implications of current trends for out-of-pocket (OOP) financial burdens in non-elderly persons with diabetes
  • High OOP costs due to increased drug spending may:
    • Burden individuals/families
    • Prevent timely health services use
      • Lead to costly long-term complications

Slide 3

Previous Literature on Financial Burdens for Health Care

• Bernard et al. (2006) found that adults with diabetes:
  • Had greater risks of high burdens
  • Had lower incomes
  • Paid a higher share of total expenditures out-of-pocket
• Banthin and Bernard (2006) found that between 1996 and 2003:
  • Prevalence of high financial burdens increased for the entire the U.S. population and within several subgroups

Slide 4

Objective

• Part I
  • Examine trends in anti-diabetic medication use and expenditures
  • Describe characteristics of persons with diabetes
• Part II
  • Examine OOP financial burdens among non-elderly persons with diabetes
    • The Elderly (≥ 65 years) are different
      • Have Medicare and mostly unemployed
• To help inform targeted policies for non-elderly persons, at increased risk of high burdens

Slide 5

Data

• Use MEPS for the years 1997 through 2007
  • Nationally representative data—U.S. civilian non-institutionalized population
  • Part I—years 1997 through 2007
  • Part II—first year of 2005-2007 panels
• Detailed information on drug purchases including:
  • Therapeutic classifications and quantity purchased
  • Source of Payment:
    • OOP payments by families
    • Private and public insurance payments
• Detailed information on health conditions, economic and socio-demographic variables

Slide 6

Defining and Measuring OOP Burden using MEPS

OOP financial burden
             =
Total family OOP spending
Total family income

• Categorize OOP financial burdens using thresholds
  • Individuals have high burden if family OOP ≥ 10% of family income

Slide 7

Part I

Slide 8

Increase in treated prevalence of diabetes, 1997 and 2007

Bar chart data:

1997: 3.6%
2007: 6.3%**

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 1997 and 2007.
** Difference in 1997 and 2007 estimates significant at p < .05.

Slide 9

Increase in comorbid conditions among persons with treatment for diabetes, 1997 and 2007

	1997	2007
Cardiovascular disease	13.1	17.2**
Hypertension	46.1	64.8**
Hyperlipidemia	15	52.8**

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 1997 and 2007.
** Difference in 1997 and 2007 estimates significant at p < .05.

Slide 10

Changes in diabetes treatment, 1997 and 2007

	1997	2007
Orals	59.9	77.3**
Insulin	38.2	24.4**
Non-insulin Injections		4.1

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 1997 and 2007.
** Difference in 1997 and 2007 estimates significant at p < .05.

Slide 11

Changes in use of oral medication classes, 1997 and 2007

	1997	2007
Sulfonylureas	51.2	40.2**
Biguanides	21.2	55.2**
Thiazolidinediones	4.7	24.6**

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 1997 and 2007.
** Difference in 1997 and 2007 estimates significant at p < .05.

Slide 12

Change in total expenditures per user, 1997-1998 and 2006-2007

	1997-1998	2006-2007
All therapeutic classes	500	944**
Sulfonylureas	298	211**
Biguanides	381	297**
Thiazolidinediones	1,013	1,121**
Insulin	331	918**
Non-insulin injectables		1,297
Oral combinations		577

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 1997 and 2007.
** Difference in 1997 and 2007 estimates significant at p < .05.

Slide 13

Change in out-of-pocket expenditures per user, 1997-1998 and 2006-2007

	1997-1998	2006-2007
All therapeutic classes	221	273**
Sulfonylureas	167	89**
Biguanides	174	116**
Thiazolidinediones	190	250**
Insulin	132	257**
Non-insulin injectables		258
Oral combinations		179

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 1997-1998 and 2006-2007.
** Difference in 1997-1998 and 2006-2007 estimates significant at p < .05

Slide 14

Percentage of persons with treatment for diabetes by sex and age, 2006-2007

Overall: 6.1
Male: 6.0
Female: 6.1
Age 0 to 17: 0.2
Age 18 to 44: 2.0
Age 45 to 64: 11
65 and over: 19.4**

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 2006-2007.
(**), [*] Differences from the reference category significant at (p < 05) and [p < 10]

Slide 15

Percentage of persons with treatment for diabetes by health insurance status and race-ethnicity, 2006-2007

Overall: 6.1

Health insurance status: Less than 65: 0
Any private: 4.0
Public only: 5.5**
Uninsured: 3.1
65 and over: 0
Medicare only: 19.7
Medicare and private: 17.9
Medicare and other public: 26.9**

Race-ethnicity:
non-Hispanic White: 6.1
non-Hispanic Black: 7.6**
non-Hispanic Other: 5.3
Hispanic: 5.3

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 2006-2007.
(**), [*] Differences from the reference category significant at (p < 05) and [p < 10]

Slide 16

Percentage of persons with treatment for diabetes by education and income, 2006-2007

Overall: 6.1

Education
Less than high school: 9.1**
High school: 6.9
At least some college: 4.5

Income
Poor/negative: 6.9
Near poor: 8.5**
Low income: 7.1
Middle income: 5.8
High income: 5.4

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 2006-2007.
(**), [*] Differences from the reference category significant at (p < 05) and [p < 10]

Slide 17

Summary—Trends in the Pharmaceutical Treatment of Diabetes, 1997 to 2007

• Proportion using major classes
  • Increase in oral use
  • Decrease in insulin use
  • Introduction of non-insulin injectables
• Proportion using specific classes of orals:
  • Decreased use of (older) sulfonylureas
  • Increased use of (newer) biguanides and thiazolidinediones (TZDs)
• Persons most likely to report treatment for diabetes
  • Those 65 years and older
  • Those with public insurance (< 65 years) and Medicare-Medicaid (≥ 65 years)
  • Non-Hispanic Blacks, the near poor, those with less than high school education and

Slide 18

Part II

Slide 19

Out-of-pocket financial burdens among non-elderly persons with treatment for diabetes by sex and age, 2005-2007

Overall: 24.3

Sex:
Men: 21.1
Women: 27.2**

Age in years:
0 to 34: 14.6
35 to 49: 20.6**
50 to 64: 27.1**

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, first year of 2005-2007 panel data.
(**), [*] Differences from the reference category significant at (p < .05) and [p < .10]

Slide 20

Out-of-pocket financial burdens among non-elderly persons with treatment for diabetes by health insurance status and race-ethnicity, 2005-2007

Overall: 24.3

Health insurance status: Any private: 17.9
Public only: 34.1**
Uninsured: 44.21**

Race-ethnicity:
non-Hispanic White/Other: 22.9
non-Hispanic Black: 25.9
Hispanic: 28.5**

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, first year of 2005-2007 panel data.
(**), [*] Differences from the reference category significant at (p < .05) and [p < .10]

Slide 21

Out-of-pocket financial burdens among non-elderly persons with treatment for diabetes by education and income, 2005-2007

Overall: 24.3

Education
Less than high school: 35.7**
High school: 23.9
At least some college: 18.0

Income
Poor: 45.8**
Near Poor/Low income: 32.7
Middle income: 25.0
High income: 9.7

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, first year of 2005-2007 panel data.
(**), [*] Differences from the reference category significant at (p < .05) and [p < .10]

Slide 22

Out-of-pocket financial burdens among non-elderly persons with treatment for diabetes by type of treatment and comorbid conditions, 2005-2007

Overall: 24.3

Type of Treatment
No anti-diabetic medications: 22.7
Injectables and orals: 37.3**
Injectables-only: 24.8
One oral class: 18.4**
Two or more oral classes: 24.5

Comorbid conditions
No comorbid condition: 17.9
Any comorbid condition: 27.5**

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, first year of 2005-2007 panel data.
(**), [*] Differences from the reference category significant at (p < .05) and [p < .10]

Slide 23

Summary—OOP financial burden in the non-elderly diabetes population, 2005-2007

• Non-elderly diabetes patients most likely to experience high burden:
  • Were: ages 50-64, women, uninsured, poor and Hispanics
  • Had less than a high school education
  • Had one or more co-morbidities
  • Used both insulin/injectables and oral medications

Slide 24

Logit: Dependent variable out-of-pocket financial burdens = 10%

Independent Variables	Odds Ratios
Age in years
0-34	1.00
35-49	1.69
50 to 64	2.65**
Sex
Male	1.00
Female	1.27*
Race/ethnicity
non Hispanic White/other	1.00
non-Hispanic Black	0.73*
Hispanic	0.79
Health insurance status
Any private	1.00
Public only	0.95
Uninsured	2.61**
Income
Poor	7.82**
Near Poor/Low income	4.23**
Middle income	3.00**
High income	1.00
Education
Less than high school	1.42*
High school	1.05*
At least some college	1.00
Comorbid conditions
No comorbid condition	1.00
Any comorbid condition	1.51**
Type of Treatment
No anti-diabetic medications	1.00
Injectables and orals	1.65*

Source: Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, 2005-2007.
(**), [*] Differences from the reference category significant at (p < .05) and [p < .10]

Slide 25

Summary of Multivariate Analysis: OOP financial burden in the non-elderly diabetes population, 2005-2007

• Non-elderly persons with higher odds of high burdens:
  • Those ages 50-64, women, uninsured, poor
  • Had less than a high school education
  • Had one or more co-morbidities
  • Used both insulin/injectables and oral medications
• Non-elderly persons with lower odds of high burdens:
  • Were non-Hispanic blacks
  • Used one oral class of medication
  • Had high incomes

Slide 26

Conclusions

• Part I
  • We highlight:
    • Growing shift away from sulfonylureas towards the use biguanides and thiazolidinediones (TZDs)
    • Swift diffusion of newer and relatively expensive classes of anti-diabetic drugs
  • We describe characteristics of persons with diabetes
• Part II
  • We shed light on:
    • Subpopulations of non-elderly more likely to experience OOP financial burdens

Slide 27

Conclusions (cont'd)

• Diabetes prevalence falls disproportionately on the vulnerable in society
  • Imposes high financial burden for health care
• The Affordable Healthcare Act may help alleviate some of the financial burden
• Our study will help inform targeted policies and additional subsidies for non-elderly persons
  • Some may still have difficulties paying for necessary medical care—even with insurance coverage

Slide 28

Limitations and Future Research

• Limitations:
  • We did not distinguish the different types of diabetes
  • Our results are descriptive in nature
• Future research:
  • Estimate a causal model