Using Medicare and Medicaid data for uninsured to price shop medical care

I have some experience with these datasets and this topic (and anecdotal experience with physicians) and I don't think this will provide anything useful. Essentially, physicians are still incentivized to bill the code that provides them with the largest reimbursement. And, the physicians are incentivized to bill for any possible code they can. What I think the dataset will provide, then, is a good look at negotiations between the physician and the government--not "what will it cost to heal me?"

In addition, health centers may very well have a policy to simply charge uninsured more for every procedure. That may be because they "negotiate with insurers for better rates" or "know that the uninsured are more risky and won't pay, so we should charge more."

The "best" thing to do, unfortunately, is call around. For a recent cut to my head, I had to call a dozen places to find the best price. I was "insured" but hadn't met my deductible yet so I'd be paying cash. Ranged anywhere from a quoted price of $850 to $120. Pays to call around.

Calling around, of course, won't be helpful if the agent misquotes you, your insurance company screws up your paperwork, or it's actually out of network.

If this is the billed data I think it is, then this data is super unhelpful. I have done a few different projects for providers (I am an actuary) and I could never find a good use for this data. Billed charges are a made up number, you really need the discounted data (varies between each payor and provider), and need to know what market you are in (employer, ind, Medicare, Medicaid, some kind of ACO or other agreement).

Also, a lot of what the reddit OP wants already exists in various forms (some of them you have to pay for, such as truven data). 'Bundled payments' have been picking up steam for the last five years or so, and there are public data sets that describe things like knee surgery costs in different geographic areas. I think CMS has a website that already visualizes this data as well.

I remember some years ago when I had surgery, the surgery was pre-authorized. After the surgery I received a bill for $20k. Insurance refused to pay the surgery place and eventually both parties settled it. I ended up speaking to someone at the insurance company that I went to school with. She said that for this procedure and amount of time for operating room the max they pay is $2k. So, there was no way that company would see $20k.

I also refused to pay as patient is responsible. Few years later I received a proper bill for $300. It was long after they could collect, so I told them to beat it.

I've been trying to figure out if there's an alternate ... game that could be played in order to find actual pricing for procedures ex post facto. My original (and I know this doesn't work) thought was some sort of competitive billing procedure. It would work like this:

1. After a procedure K is performed & the payee sends a request to billing for $M;

2. The payer can request "co-bids" from other sources;

3. The co-bidder "X" that bids the least/best gets the money (say, $N < $M);

4. But! "X" now guarantees to accept $N for that procedure K for the next (or some single future) procedure K they perform.

Obviously, a co-bidder could bid $N for procedure K with no intention to ever perform procedure K again, amongst other nefarious things.

Sounds like a way to translate the billing codes to normal vernacular would be really useful for them.