Hobservations on the RAND Research Report “Prices Paid to Hospitals by Private Health Plans Are High Relative to Medicare and Vary Widely”

1. The RAND study is a an important milestone. It reflects true Medicare re-pricing of actual commercial/private/non-Medicare patient claims on a significant number of claims, although a limited sample size.
   
   The numerical results were a little disappointing to me in that they showed a lower commercial cost amount when compared to Medicare than I believe is really the case. However, the authors delineate any number of caveats and limitations to the study probably explaining the differences between the study results and what many of us who study TPA claim repricing to Medicare believe the ratios to be in practice.
   
   
2. Generally, CMS data available for the nation, apart from Maryland (See #8 below.), indicates that the ratio of hospital billed charges to Medicare allowed tends to a range of 450-500% for inpatient, and something north of 600% for outpatient (which includes the ER). However, this data is evaluated for Medicare patients only. Hence, it has been very important to be able to analyze actual private plan patient claims from both the billed, allowed, and Medicare-allowed standpoint to adjust for any natural age bias that might exist between Medicare patients, and under-65 private plan patients.
   
   Unfortunately, one of the limits to the RAND study is that they do not show the original billed amounts, only the payer “allowed” amounts, and then the simulated “what Medicare would have paid” amounts, and so we don’t know where the claim “started” as a bill before private payer repricing.
3. It is important to take this report with a heavy shaker of salt. There is much good research, but incomplete. The devil is always in the details. The report itself isn’t overly long, has some interesting sidebars, and is worth the effort to read or at least skim, which is my recommendation.
4. In no surprise at all, Private Health Plans pay more than Medicare, with an average report ratio of 241% for the study overall for 2017. It is important to note that if this percentage is bifurcated into inpatient versus outpatient services, the two corresponding ratios are 204% and 293%, respectively. Given the generally accepted actuarial mix of charges at around 50/50 for inpatient versus outpatient facilities, these numbers make sense since the average of the two is 243%, very near the 241% overall figure.
5. The surprise result, in my view, is that this 204% figure for inpatient services is low to many of us in the industry. The geographic and other limitations of the sample set used in the study could easily reflect a bias towards states or situations reflecting lower relationships than the types of claims we see in our everyday experience. For example, many hospitals with claims in the study did not generate enough inpatient claim counts to be included, though their outpatient claims were counted. We don’t know the bias that such exclusions might have generated, though the impact was likely small if any based on the overall counts by category. The report also indicates a trend of a couple of points per year, and so 2019 ratios could be up 3-4 additional points in the ratio compared with the 2017 ratios.
6. The next most important part of the study after the findings is the section on Limitations found on report pages 14-17. That is, it is as important to understand what the report does not include as much as it is to understand what it does, and how that might tweak the results one way or the other. I believe the authors do a very good job in delineating a number of these caveats. When evaluating the relationship between private plan allowed amounts and what Medicare “would have paid,” we understand the important variations that can exist between inpatient and outpatient, between hospitals, between states, between networks, and all the combinations thereof.
7. Why is the RAND report particularly timely in today’s political environment? There is a plethora of health plan proposals from which to choose. Almost every candidate has one. Many seem revolve around some variation of the now buzz-phrase “Medicare for All.” Whether it is Medicare for some, more, all, it gets used in a manner that is largely one of pandering to some subset of voters, and almost never with confidence that the proposer has the vaguest idea of what they mean when they say it. Don’t get me wrong, some proposals come with a lot of detail, figures, estimates, projections, etc., to try to impute a well thought and conceptualized plan.
   
   Each candidate proposing a "Medicare for X" of similar single payer plan start with an unquestioned premise that is based on an unanswered question: "Is what Medicare pays providers (and has for the last 25 years) appropriate, reasonable, fair, and adequate as an amount for medical services?" This question is one for which the process of asking and answering form a gateway through which any such proposal must pass. This single question deserves a simple "yes" or "no" answer.
8. Earlier I made mention of Maryland. One of the relevant bits of what we know about Maryland can be related to a hidden question lying behind the RAND report: "What is the appropriate 'equilibrium' point where hospitals would receive the same 'allowed amount' from both Medicare and private plans, and the result would more-or-less “work?” Well, the relationship of what all payers in Maryland (or did until recently) pay as a percentage of the “standard” Medicare that RAND used for simulating Medicare allowed payments is apparently around 140%. That is, if all payers, including the government, paid 140% or so of today’s standard Medicare based allowed amount formulae, that amount would enable hospitals to survive and we could have a transparent and (importantly) non-discriminatory pricing structure. The result would save an immense amount in administrative costs, mental and emotional distress to patients and their advocates, and I assure you, actuaries.
   
   Many of you will know that I have pounded away on the Maryland “all payer” hospital system concept for many years since the state instituted the system. It doesn’t look today exactly like it did prior to the ACA, but the principles are sort of still there.
9. I recommend that you download the detailed data Excel spreadsheet and have a look at the hospital/state/network specific data. (Click here for RAND Study downloads.) The Table 2. States tab is particularly informative as to the makeup of the study data. One can see, for example, that the Colorado input was about half of the entire private plan payments in the study. It is also very easy to see the comparison between the overall weighted average ratio for a state, and the split between inpatient and outpatient ratios. There are many other findings that organizations will find of interest, and possible (though I suspect limited) use.