The Price is Too Damn High: A Multi-Part Analysis of Medicare Claims
Healthcare prices in the United States are notoriously confusing and increasingly…too damn high (in homage to Jimmy McMillan). To put it into perspective, in 2018 the Health Care Cost Institute (HCCI) performed an analysis of healthcare costs vs. healthcare utilization for individuals with employer-sponsored insurance. HCCI found that between 2014 and 2018, healthcare spending rose by 18.4% for this subgroup. Three-quarters of that increase could be explained by a rise in healthcare service prices.
HCCI’s analysis is both very impressive and very difficult to replicate, though perhaps not for the reasons you might think. The biggest stumbling block to performing such an analysis is probably the fact that private/employer-sponsored insurance claims and spending data are very carefully guarded from the public eye. To the intrepid healthcare data scavenger, the few robust, accessible healthcare spending datasets are available through the Centers for Medicare & Medicaid Services.
Ok great, you’ve now found a dataset consisting of millions of rows of medical claims! Now what? Well, we have to remember that this dataset represents just one segment of healthcare spending: Medicare spending. In 2019, only 14.2% of the population was covered by Medicare. Additionally, Medicare reimbursement rates for Hospital and Physician Services are notoriously lower than private insurance rates.
Ahhh, those are some big caveats! So can we still use this thing? Sure we can. Although Medicare reimbursement rates are not very indicative of how much the rest of us pay, we can still glean useful information from submitted charge amounts.
Back up. What the heck is the difference between a reimbursement rate and a submitted charge amount??? Well, this excerpt from CareCloud eloquently sums up the craziness of how we pay for healthcare:
In most industries, paying for a service or item is straightforward. You see the price, make the payment, and receive the item or service. The entire transaction takes a matter of seconds. Healthcare reimbursement is far more convoluted. The biggest difference between healthcare and other industries is that providers are paid after services are rendered.
Step one, you see your Physical Therapist (PT) for a 45 minute visit. Your PT gives you an ice pack, 15 minutes of “manual manipulation” (aka a medical-jargony massage), and teaches you a few exercises to do at home. You leave the appointment, happy and maybe a bit sore. Your PT then writes up some notes about your appointment, explaining what happened (the “procedures”), and these procedures get translated into medical billing codes. These are then sent to your insurance company as part of the “claim” for the appointment. The claim clearly lays out how much money your PT’s office expects to receive for each procedure that took place. These are the submitted charge amounts.
Note: watch out, because they sometimes expect $40 for that ice pack! No… I am totally not speaking from experience here…
Next up, your insurance company receives the claim. They might have special agreements with your PT which allow them to negotiate the submitted charge amounts down. So your claim has now been reduced. They then check to see how much of this reduced claim THEY owe and what YOU owe. Of course, what YOU owe is often much higher than you would like, but it all depends upon how high your deductible is, how much you have already spent that year, and other specifics regarding your personal insurance plan. The sum of what THEY owe and what YOU owe is essentially the reimbursement rate for the PT.
Now for Medicare, we know this reimbursement rate is MUCH lower than for private insurance. So let’s ignore it! Instead, let’s focus on the submitted charge amounts. Although not identical to what you and your insurance company actually pay (due to their fancy, negotiated lower rates), we can probably assume that these submitted charge amounts are reasonably correlated with the money that leaves your wallet.
Now we’re cooking! In the following multi-part series, I will use the CMS’s Medicare Provider Utilization and Payment Data: Outpatient datasets to address a number of topics, including:
- How do healthcare prices differ by state?
- Can these differences be explained by differences in median income?
- Does having more doctors offering a service lower the price for that service?
- How have prices changed over time?
- What about New York City’s data by zip code — do demographics play a role in price and/or the number of providers?
Until next time — stay healthy!