Memorandum

CDC publishes new Diabetes State Burden Toolkit – June 8, 2017

Executive Highlights

  • The CDC’s new Diabetes State Burden Toolkit allows users to find epidemiological and health economic data on diabetes in their state with just a few clicks. This tremendous resource offers figures on diabetes prevalence and incidence, hospitalizations, mortality, direct/indirect costs, and more.
  • In 2013 (the last date for which statewide data is available), diabetes prevalence in the US was 9.4% with an incidence of 7.9 per 1,000 individuals. This is up from 9.3% in 2012. There were 293,220 premature deaths due to diabetes, and 31% of these were categorized as CV deaths.
  • The average cost per patient with diabetes was $16,670 in 2013, which sums to >$421 billion for the entire US diabetes patient population. We’re pleased to see the inclusion of economic data alongside health burden and mortality data, though we would love in future updates to get a better understanding of how much of the cost might be driven by severe hypoglycemia, etc.

The CDC just released the Diabetes State Burden Toolkit, an interactive, state-by-state guide to how diabetes is affecting the health of US populations, healthcare costs, and mortality. This webpage houses a massive amount of epidemiological data on diabetes from 2013, collected, sorted, and analyzed. Diabetes prevalence for the nation as a whole was 9.4% in 2013, and was slightly higher in men (9.9% vs. 9% for women), up from 9.3% overall in 2012. This follows a trend of rising diabetes prevalence since 1990, as outlined in the CDC’s 2012 Diabetes Report Card. There were 7.9 new cases of diabetes per 1,000 individuals in 2013, though this measure of incidence was calculated based on people’s self-report of what a healthcare professional had told them, and thus doesn’t account for undiagnosed diabetes, which unfortunately persists. Diabetes was responsible for 53,150 hospitalizations for MI and 174,290 hospitalizations for congestive heart failure, corresponding to 5.6 and 9.4 per 1,000 individuals, respectively. There were 293,220 deaths attributed to diabetes in 2013 – 115 out of every 100,000 premature deaths. As expected, cardiovascular (CV) disease was the leading catalyst for mortality within this diabetes patient population, accounting for 31% or 90,950 of the total 292,220. End-stage renal disease came in second place, accounting for 11% or 33,001 of total diabetes-related deaths.

Very notably, the CDC released statistics on the cost of diabetes side-by-side with the prevalence and morbidity data. The average cost incurred on the healthcare system per patient with diabetes was $16,670 in 2013. When multiplied by prevalence, this sum exceeds $421 billion. Indirect costs (absenteeism, presenteeism or “working while sick,” household productivity, inability to work, and premature mortality) were just over $2.3 trillion. Direct medical costs stemming from diabetes neared $2 trillion, and this was split into ~$66 billion of diabetes-related expenses for Medicare, ~$26 billion for Medicaid, and ~$100 billion for other payers. Patients on Medicare had the highest direct medical cost, with an average of $5,847/person vs. $3,846 for Medicaid patients. This is likely driven by a combination of the older patient population enrolled in Medicare and the higher rates of reimbursement compared to Medicaid. We’d love to see further breakdown of this medical cost category. For instance, how much was spent on hospitalization for hypoglycemia? We wish this were easily coded and reported. How much can be explained by delayed diagnosis, or by a delay in seeking care due to high deductibles or high co-pays for prescription drugs? (Dialysis and heart attacks are also expensive, but there aren’t such public calls to reduce these costs.) These answers are critical as the US has become closer to moving toward value-based healthcare, and overall we’re hopeful that this tremendous resource of diabetes data will be influential in healthcare policy surrounding diabetes. These numbers should serve as a call-to-action for diabetes prevention efforts specifically, considering the all-too-high incidence rate. We’re pleased to see this investment and effort from the CDC to create the Diabetes State Burden Toolkit, and we encourage all of you to look up your state.

  • This toolkit comes on the heels of the CDC’s 2015 US mortality statistics, finalized this past December. Compared to 2014, 2015 featured statistically significant increases in mortality due to diabetes, heart disease, stroke, and kidney disease. Given the correlations between diabetes/CV disease and diabetes/kidney disease, this begged the question, how much preventable death is ultimately related to diabetes? The new toolkit offers an answer – 115 of every 100,000 deaths – and also serves as a glaring reminder that improving interventions for type 2 diabetes, prediabetes, and obesity could have a major impact on population health and life expectancy (which decreased by 0.1 years between 2014 and 2015, to 78.8 years).
  • Moreover, this data confirms that CV disease is the leading cause of death for people with diabetes, and highlights the need for cardioprotective therapies. It’s disheartening to hear that 75% of people are unaware of this fact, as reported in a recent Lilly/BI-sponsored survey, and we thus see critical value in data like this to promote health literacy and spur health policymakers to action. Lilly/BI’s SGLT-2 inhibitor Jardiance (empagliflozin) became the first diabetes therapy with a CV indication in December 2016, based on positive results from the EMPA-REG OUTCOME CVOT. We have our fingers tightly crossed that the FDA approves a similar label update for Novo Nordisk’s GLP-1 agonist Victoza (liraglutide) based on positive LEADER results, following an Advisory Committee meeting on June 20. That nearly 100,000 US patients with diabetes died of CV disease in 2013 is reason enough to get positive CV data on the drug label, so that busy patients/providers are aware of Victoza’s CV benefits and can bring down rates of CV death in the real world. This is just one example of how this valuable epidemiological data might come into play for regulatory practices and health policy.
  • Across all age groups, Alabama had the greatest reported diabetes prevalence in 2013 (12.7%), followed by Mississippi (12%), South Carolina (11.3%), West Virginia (11.2%), Tennessee (11.1%), Texas (10.9%), Louisiana (10.8%), Arkansas (10.5%), Georgia (10.4%), and Oklahoma (10.2%). For comparison, in 2012, Mississippi topped the list of greatest reported diabetes prevalence (11.7%), followed by Louisiana (11.5%), Alabama (11.1%), West Virginia (11.1%), Tennessee (10.8%), Texas (10.6%), Oklahoma (10.6%), South Carolina (10.5%), Ohio (10.4%), Arkansas (10.2%). We’re pleased to see a decrease in Ohio’s diabetes prevalence in 2013 (down to 9.2% in 2012), though Georgia’s substantial rise in prevalence is worrisome (up from 9.6% in 2012).
  • Seven of these states also appear on the top 10 list for diabetes incidence: Louisiana came in first (9.2 per 1,000 people), followed by Arkansas (8.6 per 1,000), South Carolina (8.1 per 1,000), West Virginia (7.8 per 1,000), Texas (7.5 per 1,000), Alabama (7.5 per 1,000), New Mexico (7.3 per 1,000), Florida (7.2 per 1,000), Georgia (7.1 per 1,000), and Indiana (7.1 per 1,000). Notably, data for age-adjusted incidence is missing for a handful of the 50 states, which may skew this list. No state in 2013 showed incidence <3.6 per 1,000, which was the rate reported for Minnesota.

 

-- by Payal Marathe, Helen Gao, Pearl Subramanian, and Kelly Close