Masters Project on Healthcare Needs of Diabetics
Title Page
By Paul Tanner Olsen
A master’s project submitted to the faculty of
The University of Utah
in partial fulfillment of the requirements for the degree of
Master of Science
Department of Economics
The University of Utah
May 2017
Copyright © 2017 by Paul Tanner Olsen
All Rights Reserved
Complications from the secondary symptoms of diabetes generate more than half of the costs associated with diabetes, including inpatient and emergency medical services. But their onset can be largely delayed until later in life if patients maintain their A1C levels through proper diet, regular exercise, and access to adequate and affordable healthcare and medications. Patients who receive quarterly A1C tests, as well as annual tests for neuropathy and retinopathy, are more likely to better maintain their diabetes. By having personal information about the status of their disease periodically, patients can better adjust diet and exercise, and their physician can recalibrate medications as needed.
Private insurance and Medicare are more likely to provide these tests, as opposed to Medicaid or Tricare/Champs VA. Private insurance was positively correlated with receiving neuropathy and retinopathy tests for both Type 1 and Type 2 diabetics, and Medicare was positively correlated with A1C tests for Type 2 diabetics in a regression statistics study using 2009 Medical Expenditure Panel Survey data. However, there is room for improvement. By providing a few consistent and inexpensive tests, diabetes costs can be reduced for individuals, healthcare providers and the economy.
The impoverished have had less access to healthcare, which makes them more vulnerable to develop diabetes, as well as secondary symptoms of the disease. As Congress works to make improvements on the Patient Protection and Affordable Care Act of 2010, they should preserve and expand Medicare and reduce the costs of prescriptions for all. With fewer barriers to entry, patients will be more likely to receive care and take medications as prescribed to manage the disease, and thus reduce costs overall.
This thesis is a long time coming after years of researching the 2009 MEPS data. I want to thank my family for their continued support that has made this project possible. I also thank my wife Emily for her editing and formatting contributions.
As a Type 1 diabetic myself, I have a special interest in this topic. I wish to thank endocrinologist James Chamberlain, M.D. and Amy Glenn, N.P., who provided a review of my criteria for distinguishing between Type 1 and Type 2 diabetics. I am also grateful to my thesis committee for their time and attention to this thesis topic.
Abstract
Table of Contents
List of Tables
List of Figures
1. Introduction
1.1. Link Between Poverty and Diabetes.
1.2. Type 1 Diabetes Is Unique.
1.3. The Impact of Secondary Symptoms.
1.4. Tests for Secondary Symptoms.
1.5. Affordable Medications.
1.6. Adjustments to Current Legislation.
1.7. Prior to the Affordable Care Act.
1.8. Statement of the Problem.
1.9. Hypotheses Statements.
2. Materials & Methods.
2.1 Criteria for Distinguishing Type 1 from Type 2 Diabetics.
2.2 Variables.
2.3 Descriptive Statistics.
2.4 Design of Regression Sets.
3. Results.
3.1 First Regression Set.
3.2 Second Regression Set.
3.3. Third Regression Set.
3.4 Fourth Regression Set.
3.5 Fifth Regression Set.
3.6 Evaluation of Hypotheses.
4. Discussion and Conclusion.
4.1. Discussion.
4.2. Limitations and Topics for Further Study.
4.3. Conclusion.
References.
Appendix A: Numerical Values of Figures in Descriptive Statistics Data.
Appendix B: Regression Results – T-scores and Coefficients.
Table 2.1. Criteria for Distinguishing Type 1 from Type 2 Diabetics.
Table 2.2. Number of Total and Total (Weighted) Observations in the Study.
Table 2.3. Number of Observations per Age Group.
Table 2.4. Demographic Variables.
Table 2.5. Variables that Are Indicators of Socioeconomic Status.
Table 2.6. Additional Variables.
Table 2.7. 2010 Race Data (U.S. Census)*.
Table 2.8. Percentage of diabetes cases diagnosed in 2000, by race/ethnicity.
Table 2.9. Commonly Reported Priority Conditions.
Table 2.10. Design of the First Regression Set.
Table 2.11. Design of Second Regression Set.
Table 2.12. Dependent and Independent Variables for Third Regression Set.
Table 2.13. Dependent and Independent Variables Used for Fourth Regression Set.
Table 2.14. Dependent and Independent Variables Used for Fifth Regression Set.
Table 3.1. Correlation Coefficients for Type 2 Diabetes – First Regression Set.
Table 3.2. Correlation Coefficients in Nephropathy & Retinopathy – Second Regression Set.
Table 3.3. Correlation Coefficients for Other Secondary Symptoms – Second Regression Set.
Table 3.4. Secondary Symptoms & Years with Diabetes – Third Regression Set.
Table 3.5. Years with Diabetes & Age – Third Regression Set.
Table 3.6. Insurance Type and Receiving Recommended Tests – Fourth Regression Set.
Table 3.7. Priority Conditions and Receiving Recommended Tests – Fourth Regression Set.
Table 3.8. Private Insurance and Secondary Symptoms – Fifth Regression Set.
Table 3.9. Demographics and Secondary Symptoms – Fifth Regression Set.
Figure 2.1. Execution of Criteria to Distinguish Type 1 from Type 2 Diabetics.
Figure 2.2. Number of Years with Diabetes.
Figure 2.3. Gender.
Figure 2.4. Married.
Figure 2.5.1. Race: Total Age Groups.
Figure 2.5.2. Race: Age 18-24.
Figure 2.5.3. Race: Age 25-40.
Figure 2.5.4. Race: Age 41-50.
Figure 2.5.5. Race: Age 51-64.
Figure 2.5.6. Race: Age ≥ 65.
Figure 2.6. Decimal Value of the Poverty Line*.
Figure 2.7. Years of Education.
Figure 2.8. Unable to Get Care/Financial Reason Unable.
Figure 2.9. Delayed Getting Care/Financial Reason Delayed.
Figure 2.10. Unable to Get Prescriptions/Financial Reason Unable.
Figure 2.11. Delayed in Getting Prescriptions/Financial Reason Delayed.
Figure 2.12. On Food Stamps.
Figure 2.13. Monthly Monetary Value of Food Stamps.
Figure 2.14. Family Size.
Figure 2.15. Ever insured in 2009/Uninsured in all of 2009.
Figure 2.16.1. Insurance Type – Tricare/Champ VA.
Figure 2.16.2. Insurance Type – Medicaid.
Figure 2.16.3. Insurance Type – Medicare.
Figure 2.16.4. Insurance Type – Private Insurance.
Figure 2.17. Percentage of Diabetics Treated for Priority Codes.
Figure 2.18. Percentage of Diabetics Who Received Clinical Tests for Secondary Symptoms.
Politicians, economists and medical executives have worked for years to provide an affordable and cost-effective healthcare system for all Americans. President Obama was the first U.S. President, with the help of a Democratic Congress, to pass a comprehensive healthcare bill called the Patient Protection and Affordable Care Act of 2010 (ACA), also known as Obamacare. As President Trump and a Republican Congress pursue revamps to the legislation, I offer them my research regarding the healthcare needs associated with a significant population, Type 1 and Type 2 diabetics, who constitute 29.1 million (9.3%) of the U.S. population. (CDC 2014).
In the United States, we spend $3 trillion, or 20 percent of the GDP, on healthcare, which is 2 or 3 times more than what other countries spend without getting better results (Solman 2017). The average medical expenditures among U.S. diabetic patients are 2.3 times higher than for people without diabetes (CDC 2014). Of diabetic patients, those who experience secondary symptoms of the disease incur the greatest expense – not only economically but in quality of life. With proper maintenance, these symptoms can be significantly delayed until later in life, resulting in fewer direct and indirect costs over time (Minshall et al. 2005).
Dall et al. (2009) reported that in 2007, of the $116.3 billion annual medical and pharmaceutical costs associated with diabetes in the United States, $58.3 billion (50.1%) are attributed to hospital inpatient costs. Another $7 billion (6%) can be attributed to diabetes-related emergency room visits and hospital outpatient costs. Inpatient and emergency room costs for diabetics involve medical treatment and procedures that are largely in response to uncontrolled diabetes.
1.1 Link Between Poverty and Diabetes
The metabolic disease diabetes mellitus is an expensive illness, especially for the impoverished. Unfortunately, there is a link between poverty and the onset of Type 2 diabetes. Although both Type 1 and Type 2 diabetics come from all socioeconomic groups, Type 2 diabetics are more likely to be poor in the United States, where there is a link between poverty and the consumption of cheaper sugary foods as a regular substitute to fruits and vegetables (Levine 2011), which leads to the disease over time. This phenomenon does not occur internationally, where impoverished populations have less access to sugary foods and rely on food staples such as corn, rice and beans. In addition to a poor diet, Type 2 diabetics are also less active, and genetics and other factors also play a role. The impoverished have had less access to healthcare and less education that would assist them in adjusting their diets to better meet their health needs.
1.2 Type 1 Diabetes Is Unique
It is unknown what causes Type 1 diabetes, which makes up 5-10% of all diabetics (CDC 2014). One current theory as to the cause of Type 1 diabetes is that certain identified genes cause an autoimmune response to turn on when a patient experiences a significant stress event, and this can occur at any time in their life (Atkinson, et al. 2014). Different human leukocyte antigens (HLA) class II haplotypes have been linked to Type 1 diabetes that develops in childhood (HLA DRB1*04–DQB1*0302, but also HLA DRB1*03–DQB1*0201) versus latent autoimmune diabetes (HLA- DQB1*0201/0302 [DR3/DR4]) whose onset is later in life (Cerna 2007). Others suggest that the condition is triggered by a virus or environmental factors (Eringsmark Regnéll & Lernmark 2013). Based on what researchers presently know about Type 1 diabetes, the condition develops regardless of patient lifestyle or health habits.
Type 1 diabetics, who are often diagnosed as children or young adults, have traditionally lived with the disease for a longer portion of their lives than Type 2 diabetics, who commonly experience the onset of the disease later in life. In recent years, a fraction of Type 2 diabetics has been diagnosed in the prime of life as a result of the obesity epidemic. Conversely, more cases of Type 1 diabetes with an onset later in life are being diagnosed as researchers learn more about the disease (Tao, et al. 2010). Over the course of their lifetimes, Type 1 diabetics will spend more out of pocket for medical expenses than Type 2 diabetics. As a result, it is important to track expenses of both diseases separately from each other.
Both Type 1 and Type 2 diabetes are on the rise. In the last two decades, the incidence of the disease has more than tripled, from 8 million in 1995 to 29.1 million in 2014 (CDC 2015-1). Tracking the disease and finding ways to cut costs will reduce its impact on patients with disease, healthcare providers and the economy.
1.3 The Impact of Secondary Symptoms
As the disease advances, both Type 1 and Type 2 diabetics will present with secondary symptoms, such as neuropathy, which can lead to the amputation of lower limbs, and retinopathy, which causes blindness. These conditions occur when elevated glucose levels in the circulatory system destroy capillaries over time. The prevalence of cardiovascular disease, another secondary symptom in diabetic patients, is similar to nondiabetic individuals who are 10 to 20 years older, both in cardiovascular events and in measures of atherosclerosis, or hardening and narrowing of the arteries (Polak et al. 2011). Renal failure, stroke, high cholesterol and high blood pressure are also conditions associated with diabetes. Out-of-pocket expenses associated with ambulatory care, emergency treatment, inpatient surgery, ICU stays, and hospice recovery will exasperate household budgets (Secrest et al. 2011), and the inability to work will further necessitate the need for patients to rely on disability insurance and social security. By delaying the onset of secondary symptoms until later in life, patients can provide for their families longer, purchase long term disability insurance, and prepare for the increased medical expenses as part of their retirement savings.
Access to affordable healthcare is essential to diabetics, who need regular doctor’s visits to properly monitor the disease. When co-pays, lab expenses and prescriptions become cost prohibitive to patients, they are less likely to go to the doctor and take their medications properly.
1.4 Tests for Secondary Symptoms
It is recommended that patients receive quarterly hemoglobin A1C tests, that identify a three-month average of plasma glucose concentration. A person with a fully functioning pancreas will have an A1C level below 5.7% (Mayo Clinic 2017-1). An acceptable target for a diabetic is less than 7%. An elevated A1C is a warning sign for patients to make changes to diet, exercise, and how frequently they are testing their blood sugar. If blood glucose levels are not monitored as recommended, the onset of secondary symptoms can accelerate.
Other tests that the American Diabetes Association recommends on an annual basis are a foot exam to check for signs of neuropathy and a dilated eye exam to measure the presence of retinopathy. Neuropathy tests may be performed by a family physician or endocrinologist, and retinopathy tests may be given by an optometrist or ophthalmologist. Some states, including Utah, require that diabetics provide the results of the retinopathy test to maintain an active driver’s license. Diabetics should also receive blood chemistry tests, such as a cholesterol test.
In addition to affordable doctor visits and monthly premiums, diabetics need access to affordable medications and supplies, which also make up a significant percentage of diabetes costs. Dall et al. (2009) reports that $27.7 billion (15.9%) was spent in 2007 on outpatient medications and supplies for diabetics. Although expensive, they help diabetics maintain healthy A1C levels, which will reduce overall expenses and ultimately help them live a longer and better quality life.
The United States has the highest drug prices in the world. There is presently no price control mechanism for drug prices in the United States. When one drug manufacturer puts a price at a new high level, other drug manufactures follow suite. Drug advertising, which is only allowed in one other country in the world (New Zealand), has increased 62% since 2012. In other countries such as Japan, prices must go down in the medical industry as technology gets older, but in the United States, that is not happening. For example, the MRI became available in the 1960s, but the cost for them is still in the thousands, where in Japan the cost is about $150 (Solman 2017).
The out-of-pocket costs for certain insulins have skyrocketed in the last few years, but market pressures may drive prices back down to affordable levels. For example, the cost of Sanofi SA’s Lantus insulin increased 90% in last five years (Loftus 2016-1). In response to the public outcry, competitors Eli Lilly and Boehringer Ingelheim GmbH have announced that they will soon begin to offer lower-cost versions of Lantus, which has prompted CVS Health Corp. to stop paying for Sanofi’s name brand product.
Medicare has faced several policy challenges this year. The Obama administration proposed a policy change in 2016 designed to curb Medicare spending on prescription drugs by requiring doctors to prescribe less expensive medications when possible, but the policy was never finalized (Loftus 2016-2). The Trump Administration has expressed its disapproval of pharmaceutical prices.
1.6 Adjustments to Current Legislation
The Trump Administration, after a failed bill in the House to repeal the Affordable Care Act, intends to pursue changes to the legislation through a Congressional act or through policy (Armour 2017). It has proposed to end the healthcare mandate that required Americans to enroll in an insurance policy, which would lead to an increase in premiums of an estimated 15-20% (Mathews 2017). It has also proposed to repeal the Medicaid expansion starting in 2020. More than 30 states have participated in the expansion of the joint federal and state insurance program designed for children, low-income individuals and those with disabilities, and the program has grown 16-25% since 2014 through the Affordable Care Act. Governors whose states have been participating oppose the rollback (Grant 2017).
The Trump administration has an opportunity to improve upon Obama’s legislation with the benefit of hindsight. If policy changes can provide diabetics with more focused care, affordable premiums, doctor visits and medications, it will move us closer to the goal of reducing the immediate and long-term costs of the disease.
1.7 Prior to the Affordable Care Act
This thesis utilizes the 2009 Medical Expenditure Panel Survey (MEPS; Agency for Healthcare Research and Quality 2011), a study of self-reported and doctor-reported medical data, to study the economic and medical situations of diabetics. The year 2009 provides an interesting snapshot of patients’ access to healthcare prior to the Affordable Care Act and allows us to assess in which areas we have excelled and where we still need improvement. I will compare insurance types to identify which is the most affordable and provides the most effective coverage to diabetic patients.
1.8 Statement of the Problem
In comparing private insurance, Medicaid, Medicare, and Tricare/Champ VA among diabetic patients, which type is more likely to:
1.9 Hypotheses Statements
Improving patients’ access to healthcare, despite their socioeconomic status, is one important method that can assist in delaying the onset of the secondary symptoms of diabetes which are costly for individuals, medical providers and the national economy. The following is a further description and the null and alternative hypotheses for the each of the three research questions.
H0: A patient’s insurance type has no impact on whether they receive the recommended tests.
H1: In analyzing private insurance, Medicaid, Medicare and Tricare/ Champ VA, at least one type of insurance will be statistically significant in facilitating the recommended tests.
H0: A patient’s insurance type does not impact the likelihood of secondary symptoms.
H1: In evaluating private insurance, Tricare/ Champ VA, Medicaid and Medicare, at least one type of insurance analyzed will reduce the likelihood of secondary symptoms.
