Up-front on PAMA impact, private payer pricing

At the Executive War College this year, Xifin CEO Lâle White talked about PAMA reporting and pricing, the erosion of private payer pricing, and market trends. Here is some of what she said on May 1, in the midst of the second PAMA data collection period, which ended June 30.

September 2019—I’m going to talk about our experience over the past couple of years with PAMA, which seeks to produce market pricing, and some of the lessons we’ve learned. We all know how we got here and one of the questions has always been: Did we ever have market pricing? The answer is probably not. When you’re dependent on a 30-plus-year-old fee schedule that never was revised for technology revisions, there wasn’t market pricing to begin with. And at the end of all of what we’re experiencing, we still don’t quite have it.

We saw in the latest Government Accountability Office report that the government is still interested in a price reduction program versus a market-based program. The report criticized some of the PAMA pricing calculation for not being based on the average Medicare price but instead on the national limitation amounts. The average price would have been an even further reduction. So in essence, the $670 million savings the government got versus the $390 million it projected wasn’t enough of an increase. The government thinks it should have gotten more.

We saw in the report the government’s criticism of the automated multichannel chemistry price bundling going away and the fact that it might cost the government as much as $10 billion. In essence, it didn’t actually cost them that, although they made it seem that way when they sent that report out in November. Essentially, panel ordering did not decrease any more than the single test ordering decreased and there was no unbundling by the industry.

However, it is important to note that the Medicare administrative contractors did eliminate their edits for panel coding for automated multichannel chemistries. So as an industry, we do have to be careful that we are still coding properly, but other than that, we’ll have to see what the outcome will be of the congressional discussions about how automated multichannel chemistries should be handled. We know they’ll add their edits back, but will they do anything else? That will be a question.

When the physician fee schedule was published last year, the applicable laboratory definition was broadened and it was a welcome change. The majority of revenue thresholds were altered by eliminating Medicare Part C from the calculation, which were the Medicare Advantage programs, so that threshold was broadened. At the same time, the National Provider Identifier definition was changed to also include more outreach labs. According to the CMS, that added about 43 percent more laboratories to the group of “applicable labs.”

But there are still questions and issues about how that turned out for us and I want to talk about that. In prior speeches I have given and seen, we noted that the lab industry itself, once you exclude the hospital inpatient, breaks down to about 44 percent of lab testing coming from hospital outreach, about 28 percent from the large labs, and another 28 percent from the rest of the labs. If we exclude the physician office labs and the hospital inpatient, the breakdown is large labs 28 percent, hospital outreach outpatient 29 percent, hospital outreach nonpatient 15 percent, and the rest of the labs 28 percent.

So what happened with CMS? Late regulations were issued at the end of February this year, similar to the first data collection period when the guidelines were provided after the collection period began. The guidance from CMS was twofold. One, that most outreach hospital labs would be included in the applicable labs but the data they would provide would be only the nonpatient data, not the outpatient data. That changed the complexity of this exercise for hospital labs, which, along with the American Hospital Association, had been pushing back that this exercise was too administratively burdensome for them. They said they didn’t have the financial systems in place to gather PAMA data and they were unable to identify nonpatient versus outpatient.

If we take a look at the way the lab market breaks down, before we take out and split up outpatient and nonpatient, had we included all of hospital outreach and in the exact pie percentages we talked about earlier—28 percent each for large and the rest of the labs and 29 and 15 percent for outreach outpatient and nonpatient, respectively—we would have seen an increase with PAMA of almost 3.8 percent if everyone had participated at that ratio. And we saw more than a 30 percent cut with the median pricing on the top 25 tests. It would have been a much different picture if we had full participation.

If, however, you take out the outpatient business from the hospital sector, it’s a much different picture. In one of the analyses of the outpatient hospital labs, it was found that 35 percent of outreach is nonpatient and 65 percent, the much bigger piece of it, is outpatient. So if you take out that bigger outpatient piece of the pie, you are left with a different breakdown: Large labs now represent 40 percent, the rest of the labs about 40 percent, and hospital nonpatient only 20 percent.

The rest of the labs and the non-patient hospital labs still represent more than half of the pie. The pricing is higher in those two sectors and we could make headway. If we take a weighted average only on this piece of the pie, we’d see that the overall price decrease in the first data collection period would have been about 8.3 percent if everyone had participated versus a 30-plus percent decline.

But when you’re talking about the weighted median, which is the way the pricing is calculated, that means more than half the pricing has to come from the rest of the labs and the nonpatient labs and this is a tough mixture to bring about. The inclusion of all these hospital outreach labs, if they do not provide any of their outpatient data—and I can tell you in the first data collection period, the hospital outreach labs that did participate, that had their own NPI and were part of the applicable labs, did provide both outpatient and nonpatient data—now their data will decline as well if we follow this new rule the CMS provided in late February.

The participation of all of these outreach labs may not move the median as much as we would like to see. Full participation will move it slightly but it may not be enough. And the CMS has already told us that if it doesn’t see enough of a movement or it makes a difference with the inclusion of outreach labs, CMS may exclude them for the next data set simply because it is administratively burdensome. That leaves us with the question about how much impact these labs will make.

Since PAMA pricing went into effect at the beginning of 2018, we have seen an erosion of private payer pricing along with it. The impact for 2018 in total was not great. Hospitals saw about a 3.5 percent decline in the private commercial business, and independent labs about 2.9 percent, but that’s because the erosion didn’t occur all on day one; it happened over the year with the majority of it coming in Q4. It happened as contracts expired, as some of the contracts allowed fees to be changed in mid-year, and when new people came into the programs.

We saw that Aetna, Cigna, the Blues, and United Healthcare started offering at the beginning of the year 20 to 25 percent off of the 2018 Medicare fee schedule, which already had a 10 percent cut in it, and it meant that a lot of these private payers were going after the full PAMA cut up front, early on, and that’s actually what they did. We saw Multiplan, which is a plan that contracts for multiple insurance carriers at the same time, cut fees in some regions by as much as 50 percent. And over time the number of individuals who go into high-deductible health plans has increased to almost 50 percent of beneficiaries. That’s another hurdle for us in the private payer market where it’s harder and harder to collect coinsurance and deductibles.

Let’s consider the top six revenue-generating tests that are high volume to take a look at what happened in commercial pricing. The commercial payers from 2017 to Q4 of 2018 cut their pricing by about 10 percent overall, and the prices that we had at the end of Q4 of 2018 (there were more cuts in Q1 of this year) were not only below the 2020 Medicare fee schedule but also below the PAMA median even from the last data collection period. So we have seen a significant erosion of pricing on new contracts by payers.

During the last PAMA exercise the molecular tests did fairly well because they for the most part had market pricing because they weren’t on the 30-year-old fee schedules. Pain management too did very well with drug testing codes. The reason for that was after the PAMA collection period, the G code definition was changed in the procedure code manual and, accordingly, the data were irrelevant so there wasn’t much change to the toxicology drug testing codes.

However, in the private sector, with the data we collected during the first data collection period, we saw a 25 to 30 percent decline over the Medicare fee for the G codes. So in the current PAMA collection period we could anticipate that the G codes will take a hit.
For molecular testing, in the private sector infectious disease pricing has gone down about 3.6 percent, in somatic testing about 2.1 percent, and in prenatal about 1.2 percent. So we’re seeing an erosion in more of the routine type of molecular testing services as well.

If we look across all payer types we see that in year one, 2018, PAMA had an approximately 1.8 percent impact across the board on revenue, and the commercial insurers, if you annualize the cuts they made in 2018, cut almost 5.56 percent off the bottom line for labs. So, again, now we’re looking at a 7.36 percent decline.

At the beginning of 2019 we saw an additional 1.6 percent PAMA cut. For the commercial payers, it was another 3.1 percent in cuts in Q1 of 2019 annualized. So pricing is eroding slowly and affecting the bottom line of a test. We’re ending up with 12.1 percent less on the commercial side, so on $100 we’re collecting $12 less.