In the context of recent reports that Medicare abruptly halted release of much-anticipated data which could have shed light on the performance of Medicare Advantage plans, CMS recently made available its “public use file” on “Provider Utilization and Payment Data” for 2015. Data provide a resource to better understand individual and aggregate delivery of health services by Medicare providers to beneficiaries, and a unique opportunity to look at delivery of mental health services in a very large population.
One aspect that has become of heightened interest to mental health providers that has implications both for provider “pocketbook” concerns as well as clinical “standards” of care is the units of time in which services are delivered, such as in 45 minute, 60 minute, or other session lengths. For example, this year, there is a 50% differential in reimbursement for a 53 minute or longer session relative to a 45 minute session ($128.12 versus $85.42 for the Medicare national payment amount in a non-facility [i.e., office] setting). Medicare is often the model for both payment amounts and “medically necessary” services.
As I previously wrote, in 2013, the first year reimbursement became available specifically for 53+ minute sessions (the “90837” code by “Current Procedural Terminology”), some mental health providers became concerned this level of service provision would not be reimbursed readily by Medicare, or by other payers. Indeed there were reports other payers were specifically limiting the number of such sessions allowed, or not reimbursing these altogether. In addition to the financial implications, these changes had the potential to affect clinical practice, with many clinicians reporting 60 minutes to be their standard session length. I previously showed that, though 45 minute session lengths were the most frequent session length provided to the Medicare population, 60 minute length visits certainly continued to be standard practice.
Now we have data for the first three years during which 90837s have been available. What has happened to use of this code, to provision of psychotherapy to the Medicare population during those years?
In 2015, Medicare providers delivered nearly 3.2 million 90837s. This was an increase of 18.1% from 2014, and an increase of 48.4% over 2013, the first year the 90837 became available.
LCSWs also showed the greatest increase in the number of 90837s provided from 2013 to 2015, a 57% increase. Psychologists showed the second greatest increase from 2013 to 2015, a 51% increase. Apparently other providers also are increasingly providing psychotherapy in 53+ minute increments, with a 27% increase from 2013 to 2015. Psychiatry showed the lowest increase among these group, with 5% more 90837s in 2015 relative to 2013.
Our information thus far:
- The number of 90837s provided to Medicare beneficiaries increased by 48% from 2013 to 2015, from about 2.1 million sessions to about 3.2 million sessions.
- Estimated Medicare payments for 90837 nearly doubled (up by 97%) from 2013 to 2015, from about $134 million to about $264.4 million.
- The increase in estimated payments is not accounted for by the approximate 8% increase in the national reimbursement amount for 90837 from 2013 to 2015.
- The increase in estimated payments is largely attributable to the increase in the utilization of this procedure code.
- LCSWs showed the greatest increase in use of this procedure code from 2013 to 2015, providing 1.5 million of these sessions in 2015, an increase of 57% from the 966,000 sessions provided in 2013.
- Psychologists showed the second greatest increase in use of the 90837 code, providing 1.3 million of these sessions in 2015, up by 51% over the nearly 892,000 such sessions provided in 2013.
While the data do not include information about possible payment denials for 90837s, this therapy session length most definitely is being reimbursed, increasingly, by Medicare. Parallel data are not readily available from for-profit payers.
As previously reported, while psychotherapy in 53+ minute increments (typically therapists provide these as 60 minute sessions) certainly was common way to provide psychotherapy to the Medicare population, psychotherapy in 45 minute sessions (90834) remained the most common in 2013. Did this continue to be the case in succeeding years? While 53+ minute sessions increased from 2013 to 2015, what was happening to 45 minute sessions? Was psychotherapy overall increasingly being provided in both session formats, or were other changes occurring?
Check back for the follow up analysis!