CMS Medication Spending and Analysis

Posted on Aug 4, 2023

Background

Diabetes is the most expensive chronic condition affecting our country. According to the CDC, 37 million Americans have diabetes and 96 million (1 in every 3 Americans) have prediabetes and  have an increased risk for developing diabetes. This chronic disease increases risk for other complications including but not limited to heart disease, chronic kidney disease, retinopathy, and neuropathy. 

The health issues come with a hefty price tag. In 2017, about $1 out of every $4 in US health care costs was spent on caring for people with diabetes. At least $237 billion is spent annually on direct medical costs and another $90 billion on reduced productivity. With this in mind, there are two key questions to answer: 

  1. How much does Medicaid and Medicare spend on diabetes medications?
  2. How can spending be reduced?

Data

To understand how diabetes is straining our healthcare system, I obtained datasets from Centers for Medicare and Medicaid Services (CMS) spanning 2017-2021. Because diabetes is growing, new medications are constantly under development and prescribed. For example, Wegovy was FDA cleared in 2021 for weight management, but the same chemical compound, semaglutide, has been used to treat diabetes in a different medication called Ozempic. Complete data for some of these medications may not be present because the industry is constantly evolving.

In order to understand how diabetes is impacting medication spending, I isolated the diabetes medications. The datasets did not classify medications by their pharmacy class, so I created a dictionary with pharmacy class (key) and medication name (value). I created a function to match the dictionary value with the medication’s generic name in order to extract the pharmacy class and create a new column. From there, I analyzed the medications by their pharmacological classifications. I focused mostly on GLP-1, SGLT2i and insulins since these medications are the most expensive and, therefore,  have the greatest potential for savings.

GLP-1s are utilized to treat diabetes and obesity by increasing insulin secretion after consuming carbohydrates without the side effects of weight gain and hypoglycemia. It delays gastric emptying and promotes satiety and therefore helps lower A1C and weight. They are often prescribed because some have secondary cardiovascular benefits. They come in daily injectable, weekly injectable or daily pill form.  In recent years, these medications have become extremely popular after they went viral on TikTok.

SGLT2is also lowers A1C by releasing excessive glucose via urinary output and carries relatively low hypoglycemia risk. These medications are also often prescribed because many have secondary renal and cardiovascular benefits. They come in a daily pill form.

Insulin has been around since the 1920s and lowers the blood sugar levels the fastest. They come in varying concentrations, durations, peaks/peakless, containers and onsets. They mostly come in injectable form. They can be taken multiple times per day, which may increase risk for hypoglycemia and weight gain.

Analysis

In terms of total Medicaid and Medicare spending, GLP-1 receptor agonists, SGLT2 inhibitors and insulins are the most expensive categorically, and their costs increase annually. As seen in the heatmaps below, GLP-1 receptor agonists are the most expensive and CMS spend billions each year on this class alone. Medicare typically covers individuals who are at least 65 years old, and Medicaid covers people with limited income. It is, therefore, unsurprising that Medicare diabetes spending exceeds Medicaid diabetes spending as diabetes is a chronic disease whose rates increase with age.

According to CMS, the most prescribed type 2 diabetes medication is metformin, a biguanide. Effective, affordable and easily taken in pill form, it’s been in use since the 1950s. It is often used as the first line of medication when a person is newly diagnosed with type 2 diabetes. As seen in the bar graphs below, biguanides take a strong lead in total overall claims, followed by sulfonylureas and insulins, which are also effective and have been around for decades. For both Medicaid and Medicare, the GLP-1 and SGLT2i have been growing in popularity, and more doctors are now prescribing these medications.

Compared to total CMS spending, the percentage of total diabetes medication spending is increasing each year. The total does not even include diabetes testing supplies and needles, so the true value of diabetes spending is even higher. At least 13.6% of  Medicare’s expenditures went to  diabetes medication. That amount increased to 17.1%  in 2017. LIkewise, Medicaid’s spend increased from 8.7% in 2017 to  10.7% in 2021.

The highest cost per claim diabetes medications include the GLP-1 and SGLT2i. Medicare pays out $800-$1600 per claim (typically 1 month supply) for GLP-1 in 2021 and $450-$1000 per claim for SGLT2i in 2021. Their rates continue to increase annually. As seen on the graphs, Steglatro cost less per claim than other SGLT2i.

How can CMS reduce spending? 

For the SGLT2i class, Steglatro/Ertugliflozin costs less than other medications in the same class per claim. Compared to other SGLT2is, Steglatro has not yet demonstrated the same positive renal and cardiac protection. However, as a diabetes medication, it still effectively lowers A1C.

This graph shows the upper limit of annual savings assuming that all SGLT2i’s can and should be replaced by Steglatro. However, complete substitution is not realistic since other SGLT2i’s have proven to be effective in reducing cardiovascular death and delaying end stage renal disease. In patients who have elevated u-microalbumin, a sign of kidney damage, or ejection fraction <40%, other SGLT2is are indicated until Steglatro can demonstrate equal protection. 

There is no head-to-head study that compares the SGLT2i effectiveness in A1C lowering as a monotherapy. It is not clear yet which SGLT2i is most effective in reducing sugars.

The graph above represents the maximum potential savings CMS can make assuming all the GLP-1 marketed for diabetes can be switched to Bydureon Bcise. In practice, Bydureon BCise has its limitations. 

When examining GLP-1 medications, Victoza, Saxenda, and Wegovy are excluded because they are marketed for weight loss. Adlyxin is excluded because it is no longer on the market as of 2023. Bydureon is a slightly harder injectable pen to use because it requires mixing (you have to shake the pen for 15 seconds). This barrier can be problematic for those with dexterity issues. However,  as a  one-time use pen, it offers the advantage of convenience, eliminating the need for measuring/dialing and the safety feature of keeping the  needle hidden until it’s used.. This injection has only one strength compared to other GLP-1 and therefore may not have the most diabetes lowering effect. This medication could be offered as part of a step therapy. If A1C goal is still not achieved after 6 months and/or patient has documented dexterity issues, other GLP-1s can be considered.

Theoretically, implementing Steglatro and Bydureon Bcise as a step therapy within their classes could potentially save CMS billions.

Future Work

The datasets have the same columns, so my codes are generally reproducible.

How can investing in nutrition and lifestyle change reduce medication spending?

How will the new insulin cap impact diabetes medication spending? 

Reference
https://www.cdc.gov/diabetes/library/spotlights/diabetes-facts-stats.html

https://www.astrazeneca.com/media-centre/press-releases/2020/farxiga-approved-in-the-us-for-the-treatment-of-heart-failure-in-patients-with-heart-failure-with-reduced-ejection-fraction.html#!

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4522877/

https://www.cdc.gov/chronicdisease/programs-impact/pop/diabetes.htm

Dataset:
https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Information-on-Prescription-Drugs

Data dictionary:
https://data.cms.gov/sites/default/files/2023-03/Medicaid%20Spending%20by%20Drug%20Data%20Dictionary%20508.pdf

About Author

Chui Pereda

I am an experienced clinical dietitian and diabetes educator looking to expand into data science. I have a growing, working knowledge of Python and R. During my free time, I like to go for a run.
View all posts by Chui Pereda >

Leave a Comment

No comments found.

View Posts by Categories


Our Recent Popular Posts


View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI