Most doctors in most settings these days are stuck picking procedure and diagnosis codes from a single sheet printed superbill with no more than 100 options. In all likelihood, it is at least three or more years behind the current code set release. Providers already have their hands full keeping track of code sets including CPT, ICD, NDC, HCPCS, etc. In four years, federal law will take effect requiring transfer from ICD-9 to ICD-10, which will expand the total pool of diagnosis codes from around 17,000 to 155,000. We will save you the nosedive into the other additions to the acronym soup of RXNORM, SNOMED, LOINC, crosswalks, crosstabs, and more.
Somewhere after the printed superbill is completed, a billing service or the practice’s own billers are then revising those selections before they go out to payers. In our experience, this revision is rarely for the purposes of medical clarity. Rather, it is an arms race to optimize revenue (within the bounds of the rules and law) against the ever-dwindling reimbursements for common codes. Inexplicably to me, there seems to be an inverse relationship between how frequently codes are used and their reimbursement. The absolute worst example of this is with the standard visit codes, I can’t think of a provider that isn’t losing money on their 992XX procedures. There is some evidence that shows payers are gaming reimbursements using clumsy statistical analysis of frequency and across-the-board percentage reductions. This is where the actuarial world of insurance, the bottom line for payer stockholders, and the day-to-day operations of medical sites meet head on. What value is the improvement to a much better conceived code set like ICD-10 if all it does is add complexity to the arms race to optimize payments? That arms race will just cause the bettors to keep asking for more and more complex “documentation” of what happened. Billers, meanwhile, will just keep coming up with more and more creative “documentation” to code the same activities that ICD-9 could cover reasonably well if it was used properly to begin with. The real benefits for ICD-10 are supposed to be an improvement on visibility of disease, improved decision support, and additional fairness in how claims are paid, not just a new weapon.
No doubt medical code sets are essential tools in practices, but without tremendously expanded training and education — much more than can be done in just the next few years — and a comprehensive existing base of already digitized practices sites, I fear these new codes sets will just foster a faster race to the bottom. Don’t forget that AHIP estimates it could cost $3 to $8 billion to implement the changes, another sacrifice at the altar of health IT (I’m still waiting for our miracle). History and common sense tells us this transition will not go smoothly. The power of SNOMED and ICD-10 can only be fully realized if providers are sufficiently adept at utilizing them, and we know without any doubt that a large percentage of physicians are not adept at using ICD-9. Physicians need clear explanations of the various code sets and unbiased best practices on how to use them, but where is the time and money for that?
See some media sources on this: http://www.bloomberg.com/apps/news?pid=20601103&sid=aBpdGTphsSqg&refer=us and http://www.ama-assn.org/amednews/2009/02/09/edsa0209.htm
Fixing Healthcare in 365 Days 