It’s Easy For A Payer To Spot An Outlier

How easy? The following information is based on the 2013 “Medicare Data” (Data.CMS.gov). Can you spot the outliers?

That’s right. It’s easy to see that the overwhelming majority of clinicians provided 1000 or fewer 53+ minute sessions in 2013. A few obviously break away from the pack providing 1-2,000 such sessions, fewer still between 2-3,000, and a few clearly identifiable individuals billed for more than 3,000 such visits in 2013. One clinician billed more than 8,000 of these visits.

For this example, the “Clinician Identifier” has been “de-identified.” But the first time I drew this chart the identifiers were National Provider Identifier (NPI) numbers linked directly to specific clinicians. Taken directly from available data. The data are public. The data are searchable. The data were released with the intent to increase transparency of practice. An industrious patient, a journalist, a competitor, can run the numbers. Medicare certainly will, particularly for high cost/high utilization procedures. (Lest you argue that psychotherapy surely cannot fall into such a category, consider that the 2013 practice patterns within Medicare imply in excess of $396 million in payments for psychotherapy visits.)

All payers will have parallel numbers for practitioners who submit charges to their system. Payers will run the numbers.

They have the data and they know how to use it.

If a payer such as Medicare, or their hired recovery auditors, look at this from a purely statistical basis — at least as a “first pass” to identify clinicians to audit — a few clicks in a spreadsheet or even a built in algorithm working in “real time” make the task simple.

There are several ways to look at the extent of the variation from the norm the outliers’ numbers are. For example, for the 53+ minute therapy code (90837), in the Medicare population in 2013, here are what the numbers look like.

On average, clinicians billed 187 of these sessions in 2013. One hundred sixty two clinicians billed above three standard deviations more than that, i.e, billed more of these sessions than 99.7% of their colleagues.  The 162 clinicians who billed > 3 SD above the average number of sessions billed an average of 1,446 sessions (range of 878 to 8058). Compared to an average of 187 sessions for all clinicians. That is almost 8 times as many sessions as the average clinician.

Another way to look at this is, working every one of 40 hours of all 52 weeks of the year, there are 2080 hours available for direct service. Eighteen clinicians billed more than the number of hours available in a 52 week, 40 hours per week year.

Similar comparisons are readily available for 30 and 45 minute session lengths.

On average, clinicians billed 233, 45 minute sessions. Five hundred fifty three clinicians — almost 3% of all clinicians — billed more than 3 SDs above that amount, billing an average of 1670 such sessions, seven times more than the average clinician. One hundred eighteen clinicians billed more of these sessions than there are number of hours in a 40 hour-week/52 week year.

Here is a graphical summary of the averages and “outliers.” Understand that the definition of a statistical “outlier” practice I am using in this example is a number of treatment visits more than 3 standard deviations above average. Payers easily could define outlier practice at 2 SDs or other “cutoff” values.

There may be reasonable explanations for the unusual numbers. But the numbers would easily get clinicians identified.

Physicians may be “used to” receiving feedback about their practice patterns from payers, at least in the sense of receiving the information, though perhaps not in the sense of comfortable acceptance. Mental health clinicians may be much less familiar with receiving such feedback from payers. If past history is any guide, mental health clinicians in particular will experience such “information” to be intrusive and excessive oversight into ethical and effective practice. After all, mental health has experienced decades of disparate pressures toward downward payment and excessive, intrusive scrutiny, to the extent that special legislation was required. So heightened sensitivity to scrutiny is an understandable response.

As clinicians, we certainly will be attentive to the health of clients and the health of our practices. My sense is that, as citizen-taxpayers, we ought to be as concerned about the health of the entire healthcare system. How do you feel about clinicians billing more than 7 or 8 times the number of treatment visits offered by the average clinician? Would you want such practices identified?

Payers will scrutinize costs. That’s business. They have the data and they know how to use it. Clinicians need to get used to it.

Actually, a proactive approach may be ideal: know the standards among your peers for provision of services and measure your own practice patterns. At least within the Medicare system, a clinician receiving notification that s/he billed a particular procedure at a substantially different frequency than peers can verify the data. But commercial insurers will keep their numbers secret as “proprietary and confidential.” If a clinician receives such notification from a commercial insurance company, how would they know the report is accurate?

The next step in transparency must be that such data held by commercial insurers are made public.

Leave a Reply