A change in coding for behavioral screenings illustrates how payer payment policies negate any big-data promises of ICD-10.
For eight years, we have conducted behavioral health screens, some written and some verbal, at all well exams. We do this for two reasons. First, we believe in the screens to identify otherwise undetectable behavioral health needs in children. Second is the effect of the Rosie D. lawsuit in which the State of Massachusetts lost because they were not adequately screening children for behavioral health needs. In those eight years we've found what current studies support: By identifying behavioral health needs early, interventions are less costly and more effective. In addition to the behavioral health screens we also do a developmental screening interview at well visits for younger children, up to age 3.
Here's our current issue: For seven and a half years, we administered our screens and billed them as a CPT 96110 (both developmental and behavioral screens used the same code) and got paid by insurance anywhere from $0 from payers who bundled the service to $25 on the high end. Mostly we received $9.73 as that's what our state's Medicaid program pays. You see, all was fine until January 1 of this year when NCCI edits changed the CPT used to differentiate between behavioral screening tools and developmental screening tools. This is where our current nightmare began.
Beginning Jan. 1, 2015 NCCI edits added CPT 96127 for emotional and behavioral assessments. This meant that some of the screening tools we used would still be classified as a 96110 while others would now be the new 96127 code. To accommodate this change, we made a few simple tweaks to our integrated EHR/practice management system and started using the new, correct code. It didn't take long, however, for us to realize that many payers were not recognizing 96127 as a preventative code. This meant that for some of the routine screening tools, patients were receiving copay and deductible bills. This left us, as well as pediatricians across the country, with a coding dilemma with four bad options. Let's review:
BAD OPTION #1: Bill the Patient
The insurance companies are processing the claims and their EOBs, sent to both us and to the patient, clearly state that the copay or deductible for the "behavioral health assessment" is the patient's responsibility. It doesn't matter that we, the provider, think the payer is wrong in pushing the charges onto the patient; both developmental and behavioral screening are separately defined as preventative by the inclusion of Bright Futures in the Affordable Care Act. In these matters, we are only the insurance company's collection agency. We could, and currently do, simply mail the bill to the patient informing them that the decision to charge the patient is all on their insurance plan and to contact them, not us, with questions. Of course this is a bad choice because patients we bill are angry that we would perform any service their insurance company doesn't deem preventative.
BAD OPTION #2: Play the Game
A number of our colleagues play the game of coding to get paid, as opposed to accurately coding. We understand that it is standard practice to "check with the payer" and fudge one's insurance claims to manipulate it into something that will have the biggest payoff. This bothers us though, because it’s lying. More importantly, if we fudge claims data and know others do too, it makes it hard to believe in the power of big data and the ability of things like ICD-10 to track public health concerns. In fact, the paper we cited earlier showing the effectiveness of screening tools used coding data tracking the 96110 code and specific modifiers to build their conclusions. If coding is fudged, it's not useful.
BAD OPTION #3: Stop Doing the Screens
If you missed it, we highly recommend reading Atul Gawande's most recent article in The New Yorker about the current crisis of unnecessary medical testing. Could it be that these screens are just a waste of time and money? We have heard of peers who only offer the screens to patients who have Medicaid, but aren't they risking a failure diagnosis suit by not screening everyone?
BAD OPTION#4: Do the Screens and Don't Bill
We could still perform the screens and identify health problems in our patients and just not bill for it, or we could bill the payer, but then not bill the patient as the payer instructs, but we're not going to do that. We have no interest in being martyrs and will fight to be paid for the work we do. And, while it seems kind to not mail the deductible bill, that is highly problematic. Again, the more significant issue here is that by not coding for the screens, we'd interfere with any ability for researchers to use coding data for epidemiological studies.
For now we are choosing bad option 1, but admit that we are not 100 percent confident it's the right choice. What would you do? Are there other options we haven't even considered? Tell us in the comments below.