The Next Medicaid Therapy Rate Cut and Policy Changes for 2017

Early yesterday the TX HHSC leaked a pre-announcement of its intentions to propose a series of policy changes and adjustments (cuts) to therapy rates slated to be implemented on July 1, 2017. This was done, according to TX HHSC, to comply with the next steps of Rider 50 language from HB1 passed by the 84th Texas Legislation in 2015.

This is a summary of the proposal:

  1. therapy home health untimed CPTs will be reimbursed based on timed units to match the 2017 AMA CMS CPT code designation of untimed or timed and the new unit reimbursement will be based on a “current average” billing amount,
  2. OPT/CORF and independent therapists (general rehabs) timed CPTs will be reimbursed based on untimed encounter visits to match the 2017 AMA CMS CPT code designation of untimed or timed and those CPTs will be reimbursed based on “current average” billing amounts,
  3. uniformly now, assistant therapists-coded CPT services will be reimbursed at 70% of the normal rate (Note: Superior has already been implementing this since July 2016), and
  4. the speech reevaluation code will be reimbursed at a lower rate than the initial evaluation rate for all delivery models.
These were referred to as initiatives 1 to 4. For initiative 3, published rules will be made available on March 24, 2017 and the final rate hearing for all 4 initiatives will be in May 2017, with the proposed rates to be released at least 10 days before that hearing.

However, the devil’s in the detail. First, both the home health and office setting delivery models for Texas Medicaid will match the designation from the 2017 AMA CMS CPT description as to whether they are timed or untimed. In other words, a therapy CPT code that is designated as untimed(timed) in the 2017 AMA CMS schedule will be equally designated as untimed(timed) in the proposed Medicaid rate schedule for all therapy delivery models. The largest effects here will be felt with the speech treatment code 92507, which is an untimed code in the 2017 CMS schedule.

After all of the timed/untimed adjustments are made for CPT codes, the rate changes will depend on what TX HHSC means by “average billed” in their rate adjustments. Normally, TX HHSC is not privy to up-to-date statistics until the end of a quarter or half. The most logical definition of an “average billed” should be based on Q1 2017 (because of the first effects of the last cut on the Medicaid therapy budget), but could also be based on 2016 altogether, to obtain more robust statistical estimates. However, remember that the last cut, implemented in December 2016, is only now showing its full effects.

Part of that effect was, of course, pre-implemented by the Superior discount rate cuts and policy changes of July 2016. So, it would be much more accurate statistically to use January 2015 to the end of Q1 2017 (March 31, 2017) as a basis for that averaging. We have some summary data on 2016 that Commissioner Smith presented to the Senate Finance Committee last year, but it obscures any trends in therapy reimbursement decreases.

Nonetheless, we submitted an Open Records Request to TX HHSC for the most current therapy claims data (1/1/2015 to 3/17/2017), but it could take months to get that data (if it is even available or to be made available) based on TX HHSC’s willingness to release those data, possibly noting other superfluous concerns. We have also put in a request for clarification from the rate analysis team at TX HHSC, as to what time period they will collect from and what averaging methodology will be used. There are no guarantees of anything coming from them at this point.

Once more, the proof is in the pudding – what periods of time will those normalized (averaged) rates be based on and can we rely on HHSC actuaries to give an accurate analysis of those normalizations? There are a number of ways to “average” datum and it comes down to a smeared statistical estimate anyway. For example, if data is collected from 7/1/2016 to present, then the data should reflect the full effects of lower MCO payments that were implemented due to discount rate decreases and policy tightening and the later implemented 12/15/2016 rate reductions. Compare this data to the data that would co-mingle pre-07/01/2016 data with post-07/01/2016 data. The smearing that would average out these effects would reflect a higher reimbursement than from the first example of data. The most obvious way the TX HHSC will probably compute averages for the new rates will be to simply average out all the reimbursements per encounter for each therapy CPT code, regardless of the number of units billed for that CPT code when estimating the most likely encounter-based reimbursement amount for office settings. For home settings, they will average out the same encounter reimbursement amounts and normalize by the effective number of units serviced that have to be reported, which could be 4 units, but could be less based on a sample of surveyed home health therapy treatment documentation from MCOs and the TX HHSC.

We need to know at least two things, (i) the period of time that data is being collected and used, and (ii) what methodology will be used for averaging, (i.e., for office settings, will the untimed average rates be calculated by averaging the total billed amounts per encounter visit or will it be averaged by the total billed amounts divided by the total units billed and then multiplied by 4, and for home health settings, will the average be calculated by averaging out the total billed divided by the number of encounters and then divide by some number of units less than or equal to 4).

As examples, let us say that for CPT x, and for the office setting case; arO(x) = average untimed rate for CPT x, tbO(x) = total billed amounts for CPT x, neO(x) = number of encounters for CPT x, and nuO(x) = number of units billed for CPT x, then we can have two possible scenarios – (1) arO(x) = tbO(x)/neO(x), or (2) arO(x) = [tbO(x)/nuO(x)]enu(x), where enu(x) is an indication of the “effective” number of units serviced, on average, for CPT x per encounter. For the home health setting case, we have the scenario, if arO(x) = average timed rate per unit, then arO(x) = [tbO(x)/neO(x)]/enu(x). In either case, enu(x) is the critical effective number of units that TX HHSC will be assuming in its averaging. For the home setting, the further enu(x) is from 4, the higher the average, arO(x). For the second office setting scenario, the further enu(x) is from 4, the lower the average, arO(x). So, enu(x) is key for the new rate calculations.

UPDATE: As of March 23, 2017, we have been notified by HHSC’s rate analysis team that the proposed new reimbursement rates will be based on fiscal 2015 data and the averaging methodology will be based on the averaged billed amount for each CPT claim line. With this said, we modified our open records request to HHSC to obtain fiscal 2015 data for therapy reimbursements to include the following for each delivery model and encounter: (1) date of service, (2) CPT, (3) units billed, and (4) amount billed. For home health, the units billed will almost always be 1 and so, we are assuming that HHSC will be dividing the billed totals for a CPT by the number of encounter visits and then be a number between 3 and 4 (using an assumption that the Texas A&M study utilized). In our analysis, we will compute a sliding window of reimbursement rates based on the sliding window of 2 to 4 units per home health visit assumption in order to give an indication of the possible rate changes. Additionally, we will be analyzing the therapy rate trends and compare them to other provider type rate trends in Texas and in all other states’ Medicaid programs. These visuals will act as a comparative trend analysis and study of how the legislature has disproportionately affected therapy services in Texas since 2008.

On a slightly brighter note, Rep (Dr.) Zerwas has published and entered into consideration HB 1 and into the Appropriations Committee (since 2/13/2017), the House’s version of the budget bill. In that bill on Section 17.10 – Adjustment of Therapy Rate Reductions, it is proposed to increase the budget for therapy reimbursements by $14M during the next biennial (2018-2019). What this effectively would do is increase the overall therapy budget to $118.8M for 2018-2019 from the current $104.8M that is slated. The corresponding Senate bill SB1, does not specifically contain language for any therapy cuts. Both bills must be reconciled in subcommittee before the final votes are applied in the chambers. In the end, even if HB 1 is passed and implemented, TX HHSC may not adjust the rate schedule up (if at all) until 2018, the beginning of the new biennial. Additionally, how much of that $14M increase would level off the effective rate reduction scheduled for July 1, 2017?