Adam Sacarny has a pretty awesome and interesting (to me at least) NBER paper that looks at coding practices in hospitals when there was major money on the line.
Here is the set-up. Before 2008, CMS allowed non-specific Heart Failure codes to be paid at a high level. Almost no hospitals submitted claims with detailed, specific codes as hospitals don’t get paid for more specific documentation. Starting in 2008, the non-specific codes were paid at a low level while specific codes were paid at a high level. There is no change in the underlying characteristics of patients, just a change in the documentation. The total potential revenue swing is large (2%) and fairly straightforward to capture.
hospitals clearly respond to the incentive: within weeks, they're using detailed codes enough to capture 30% of the heart failure revenue. 3 years later, they're capturing about half the revenue.
put the other way: they left a lot of $$$ on the table, too. 5/ pic.twitter.com/iNsPgF9Ljn— Adam Sacarny (@asacarny) April 9, 2018
These results surprised me a lot.
My prior was that if a claim element drives payment, it will be addressed.
I could expect some learning by doing in the first few months as old work flows are broken and new work flows are implemented. Perhaps new coders were hired? Might some docs be trained to say a bit more in their notes? Maybe there would be a proliferation of billing optimization consultants swarming over hospitals with their Powerpoints with their schemes? I would have thought that by three years, almost all (95%+) of the revenue potential would have been captured.
That was not the case. Hospitals were leaving serious money on the table due to documentation.
His table 6 showed that large hospitals in urban areas with significant teaching responsibilities were most likely to capture more revenue than other hospitals. I am speculating now, but I wonder how much of this type of hidden management/operational expertise drives consolidation as a big, urban, teaching hospital could look at a community hospital and see that there was a major documentation revenue center that could be quickly exploited as a source of “value” (from the Merger and Acquisition viewpoint) while the longer term plans of buying out referral patterns and building a hub and spoke feeder pattern would provide medium term benefits.
I found this paper fascinating as my assumption that if a code gets paid for, it will be coded has been shaken.
Tokyokie
The big urban teaching hospitals usually will have their own coders in their HIT/HIM departments. The small rural hospitals, which generally don’t have enough patients to justify hiring coders, will contract their coding to third-party companies. That’s probably where the difference is. Also, it’s not like the small rural hospitals are likely to have cath labs and the like. Most of their heart-failure cases will probably be stabilized and then shipped to a larger hospital with those capabilities.
The Ancient Randonneur
I don’t see how the capture rate approaches anywhere near 95% unless you remove coders and implement a system where the medical professional directly inputs to a database. I don’t know what the most effective system would look like but it certainly requires substantial investment in technology and a it has to be relatively “pain free” for the medical professional to use.
Wag
Tell me how you measure me, and I will tell you how I will behave.
Another Scott
@Tokyokie: I was thinking along those lines as well.
I wouldn’t have expected 90+% uptake unless the time and effort were absolutely minuscule.
Plus, physicians get set in their ways and don’t like change, and are quite often rushed. Any paperwork that takes more time for them will be resisted – even if it makes the hospital more money.
But the big takeaway for me is – this is another example that (in general) peoples’ actions follow incentives. I hope someone is thinking about whether the extra time spent on the detailed coding is actually helping improve outcomes (or lower costs). More accurate information is (in general) better, but it costs money and time and there are only so many hours in the day…
Cheers,
Scott.
Mary
A little off topic, but when my son needed speech language therapy I had to tell the provider’s office how to code it so they would get paid, and not charge it to me. They would sometimes submit two codes, a general and a specific, and if they put in the general first it wouldn’t get paid, and I would have to ask them to resubmit with the codes flipped. It’s a lot of work, and I don’t know how much incentive there is to the coder to maximize reimbursement.
nonynony
@Another Scott:
This has a lot of explanatory power, and not just for how physicians behave.
If it’s just a “directive from management” then it will be followed up to the whims of the employees as to how much they want to conform to what management tells them to do and how much room they have to keep doing the job the way they always have. Telling people to change what they’re doing when there’s so benefit to themselves has an incentive mismatch – “if I change what I’m doing I do more work for the same pay, and my employer pockets more money because of it” is not an incentive to change your processes.
If you incentivize the change at the employee level, then you can expect to see movement towards behavior change. Even there it won’t be universal because some folks prefer to just keep doing the job they’ve always done even in the face of incentives to change. This is true of any large organization, I fully expect it to be true of hospitals.
Nate Combs
What are the chances of it working the other way, and the remaining half of reimbursement was a result of initial upcoding/overpayment?
I am in a conflict with a healthcare provider currently where they billed me on a code for a current-patient-visit, where the doctor makes a 6-bullet examination and makes a low-level medical decision (MDM), which is a 99213. All I had actually done was go into the office to pick up an inert medical device (orthotic lifts), which the doctor personally handed to me and expressed well-wishes that they’d help.
Under the 99213 code I was billed $130 for that hand-off, which, if their intent was to bill me at all, should have been at most a 99211 code (office-patient interaction, at the level of a prescription-refill). It took me 9 phone calls, months of time, and threats from the office to send me to collections before they adjusted their code. Their main point, repeated to me ad nauseam, was that the doctor had seen me, so he could code whatever he liked. My insurance company approved the code as soon as it was submitted, giving a $28 adjustment, and applying the remaining $102 to my deductible. There was no information in the notes for that visit/billing code that would have come anywhere near meeting the actual billing criteria, so I am concluding that the dollar-amount was so relatively low to my insurance company that they rubber-stamped it.
If the difference between specific and non-specific coding directly correlates to billing, and the amount of money paid out by patients and insurance, I am much more likely to believe that the payers are getting fleeced.
In my business, the more line-items a budget has, the more expensive something will be, since the little details tend to get accounted for ‘officially’ instead of included bulk with the service. In a medical environment with massive amounts of high-deductible plans, I am going to believe that everybody upstream from me is going to be out to get a buck.
Bob Hertz
I am glad to see this post, because not enough attention is paid to the importance of coding for what we call the cost of health care.
Left to me, I would wipe out all the opportunity to upcode by having one low code for all episodes. It would be even better to have hospitals on annual budgets and need no coding whatsoever, but that is a reach too far.
EL
@Bob Hertz: That would be problematic. The difference between a visit for a prescription refill and a visit for a patient with heart failure, diabetes not well controlled and with many complications, etc. is very large.
In terms of the he original question, I (try) to teach physicians how to document and code correctly. Bottom line is that a lot of it is complex, confusing, and often makes no sense. Many physicians will only spend so much mental energy on something that doesn’t impact patient care. If they are busy, they may well write “heart failure” and leave it at that.
EL
@EL: argh, where did the edit function go? “Write” not “right.”
Tokyokie
@nonynony: The coders can always query the providers when documentation is missing. And yes, the physicians hate that. But as cases are supposed to be wrapped up 30 days after discharge, physicians who consistently fail to provide sufficient documentation in a timely fashion will soon have the hospital administration giving them hell and threatening to take action against them. And the possibility of loss of privileges can be a powerful persuader.
Bob Hertz
Codes certainly have their uses for research…but they should not control how a provider is paid.
I swear by prepaid doctors and hospitals, like Kaiser.
Sab
This whole thread is terrifying but enlightening.
I am just a patient, as is my spouse. His ex is a coding genius but she is his ex.
Eric
@EL: I agree that one of the main problems is that it’s too difficult. For my field I could easily code 3-4 codes per payment per outpatient visit but that would neither benefit the patient, my staff, or me. The care my patient gets is the same. It would be more work for me and have a greater chance of being flagged or denied based on documentation. So I code enough, document what I did, and don’t generally go after that 2% extra.
As it’s already been pointed out, the incentives are wrong. It’s more work for me (we don’t have codes for outpatient visits), takes more time, feels unethical to do something extra just for the money (if it’s medically necessary that’s different but usually it would be for some insurance company’s idea of what’s necessary or normal), and I wouldn’t see a dime but would occasionally have to explain to patients why their bill was so much. So heck no.
And while icd9 was probably usefull but too general for research, icd-10 is far too detailed to be useful for research. My field has literally dozens of codes that are similar and so I use just a few that mostly work but aren’t as specific as they could be because it would take too much time with no benefit to me or my patients. So I don’t trust my own codes, let alone others in my field, and definitely not codes for other fields. Do we really need one code for bit by a parrot and another for bit by a macaw?