FREQUENTLY ASKED QUESTIONS
Brand Drug List Price Change Box Score
Hello fellow drug pricing nerd, if you’ve navigated to this page in an effort to expand your knowledge of the many teachings of the drug pricing force, RESPECT… You clearly are interested in diving into the very nuanced and geeky weeds of our sacred Jedi texts to better understand how the 46brooklyn Brand Drug List Price Change Box Score works. If you haven’t done so yet, we highly recommend reading the launch report that we released in conjunction with the dashboard.
So, fellow kindred spirit, here is how this page is organized. Below you will find a section on each Stat Box within our dashboard, with answers to anticipated and frequently asked questions (FAQs) on each section. We also kick off the party with a section on how we came up with our drugmaker list, as this exercise is not as simple as it may seem.
Since you are already here, we’d urge you to at least skim through all of the sections, rather than skipping to your Stat Box of interest, as the sections build on each other. For example, the Stat Box #1 section spends a lot of time explaining how we count drug price changes, something you’ll need to know for Stat Boxes #2 and #3.
With that as fair warning, may the Force be with you as you go forth and conquer.
General
FAQs
1) With all the rebates sloshing around behind the scenes, do changes in list prices even matter?
We recognize that there are many folks out there who are not fans of drug list price analyses. In many respects, we would fall into that camp – especially when viewed in isolation. It is very true that a continually growing mountain of price concessions from brand drug manufacturers are paid out to government programs and pharmacy benefit managers (PBMs), before some portion of these dollars make their way back to employers and other payers. Those rebates – and the growing mass of associated fees that walk and talk like rebates – without question create poor incentives that often encourage manufacturers to push list prices higher. And while it’s tough for us to pin down exactly what’s happening to net prices in the aggregate, there is research that suggests that post-rebate drug prices could actually be falling (there’s also research that suggests the opposite). Regardless, prescription drug list prices are over-inflated, and as we’ve discussed previously, they’re essentially “fake.”
So, does this mean we should just totally ignore what’s going on with list prices?
Absolutely not!
In a nutshell, here’s how this works. If you get a prescription filled for a brand drug, the cost you are subjected to at the pharmacy counter is right around the list price. So, when that list price rises, your cost rises with it. Of course, this may be hidden from you by your health plan design. Say you have a $50 copay on a brand name drug. That may not change immediately just because a drug’s list price went up, but after enough upward pressure, that $50 copay may become $100. The impact to you is much more direct if you are in a high deductible plan and haven’t met your deductible. In this case, you will directly pay for the increases in list price. Or if you happen to be a member of a health plan that isn’t receiving or flowing through all those big drugmaker price concessions, increasing drug prices could mean higher premiums or eroding benefits coverage. So higher list prices – without dollar-for-dollar discounts offsets – one way or another, equate to higher patient costs. This is the way the system is designed, even for the largest prescription management program in the country – Medicare Part D is explicitly designed to subsidize 74.5% of the program’s drug expense, leaving 25.5% to the patient.
Some time after you pick up your list price-based prescription, rebate dollars start flowing back through the leaky plumbing of the drug supply chain. Some portion of these dollars make it back to payers (whether it be your employer or a Part D plan sponsor) who then (hopefully) use these dollars to subsidize premiums (or a “less worse” plan design) not for just you, but for your entire group.
See the problem here? The system is designed to slap people taking drugs with an inflated cost, and the distribute the benefits (i.e. rebates) to everyone. So how do we lower health insurance premiums in such a system? Simply create more sick people to generate more rebates. See: SCPO.
We need to think of drug list prices as the “gravity” that holds this country in this constant state of distress, frustration, and distrust when it comes to the cost of our prescription drugs. Clearly, gravity doesn’t always tether us within a few feet of the earth. Hop on airplane, or strap on one of those awesome Mandalorian jet packs, and you’ll be able to break free from the constraints it places upon you. If you are lucky enough to not take any expensive medications right now, you may not see it or feel it, but you have one of those jet packs on right now, as you are at least partially benefitting from the drug rebates cooked into your health insurance cost. Heck, if you hover in the air (i.e. remain healthy) for long enough, you may even convince yourself gravity doesn’t even exist. But disease is always lurking around the corner, and when it strikes, patients will face a hard wakeup call (or should we say, “fall”) on how real and powerful list price “gravity” truly is.
2) Normally 46brooklyn focuses on Average Wholesale Price (AWP). Why base this dashboard on Wholesale Acquisition Cost (WAC)?
Sadly, drugs have more prices than there are types of stormtroopers. We normally focus on Average Wholesale Price (AWP) because that is the benchmark predominantly used to set prices in contracts between PBMs and employers/providers. In other words, AWP is the most meaningful benchmark price when it comes to those paying the bills for drugs as well as those getting paid for drugs.
But for brand name drugs, drug manufacturers are not actually setting AWP. Rather they set the drug’s WAC and submit it to a handful of companies that maintain and publish drug pricing compendia files (e.g. MediSpan, First Databank, Elsevier Gold Standard, IBM Micromedex Red Book), who then nearly always calculate the AWP as a 20% markup to WAC. As such, with this dashboard being focused on the prices brand drug manufacturers are setting, it made more sense to use WAC than AWP.
Be warned though - this explanation only holds for brand name drugs. Generic drugs are a totally different story. For generics, Suggested Wholesale Prices (SWPs) are submitted to the aforementioned companies, where they typically become AWPs. Generic drug manufacturers also set WACs for their drugs, but they are largely meaningless for payers and providers. So if we had chosen to focus this dashboard on generic drugs, we would have used AWP over WAC.
Clear as mud?
Using the dashboard
Select Drugmaker
The first decision you are confronted with when you arrive at the Brand Drug List Price Change Box Score is which drugmaker to choose. The viz defaults to “All” drugmakers but is also designed for you to select just one drugmaker and view all the Stat Boxes for them. Note that this is a global filter. In other words, all Stat Boxes will update with information only for the selected drugmaker when you choose one from the dropdown box
But as is the case with all things drug pricing, assigning the drugmaker to a drug is not as easy as it seems. As such, we have created this FAQ for you to proactively answer some questions that we anticipate we will receive on how we assigned the drugmaker.
FAQs
1) Where do these drugmaker names come from?
All drugmaker company names come from Elsevier’s Gold Standard Drug Database (GSDD), which we have heavily leaned on to create the entire visualization. However, the company names in the dropdown list are not a direct pull of the companies Elsevier has linked to each drug. We have added a bit of processing to hopefully make them a bit more helpful.
Within GSDD, Elsevier houses a database of company affiliations. For example, there is a table that tells us that Sandoz is an affiliate of Novartis. Well-known relationships like that are easy to flag, but with thousands of labelers out there, and a mind-bending amount of consolidation amongst these labelers, we needed to rely on Elsevier’s parent-child relationships to catch the majority of them. So, we use GSDD to map each drug up to the highest level of Elsevier’s parent-child pyramid. We then manually spot-checked the resulting list and added a few of the more recent combinations that are not yet in GSDD (i.e. Abbvie and Allergan).
While our resulting list is certainly not perfect, it should be a much better representation of each drugmakers’ complete portfolio than had we just relied on labeler. If you see something we missed, drop us a note.
2) Is this field actually the manufacturer of the drug?
Technically, no… it is not. It is actually the drug’s “labeler.” According to the FDA, a labeler, “may be either a manufacturer, including a repackager or relabeler, or, for drugs subject to private labeling arrangements, the entity under whose own label or trade name the product will be distributed.” That’s about as easy to understand as Darth Maul surviving his Kenobi bisecting. A lot of the noise in this labeler definition has to do with repackagers, who do exactly as their name indicates – repackage drugs from the source drugmaker under their “label” and market it under a new NDC.
The good news here is that our entire dashboard filters out NDCs marketed by repackagers, so we don’t have to worry about them. But that still doesn’t mean the labeler is the manufacturer. It could be, “an entity under whose own label or trade name the product will be distributed.” This is when the manufacturer agrees to produce a drug for another “drugmaker,” who then markets the drug. A great example of this is with authorized generics (AGs), where AG labelers simply market brand name drugs manufactured by brand drugmakers, with their own label and NDC slapped on the bottle.
More good news – there are no AGs in this dashboard. Its intent is simply to measure price increases on the truest of true “brand name drugs,” which we define as those that are considered “Brand” drugs by Elsevier Gold Standard Drug Database (GSDD) and that were approved by the FDA under either a New Drug Application (“NDA”) or Biologics License Application (“BLA”). Typically for these drugs, the labeler reported to the FDA is the drug’s manufacturer … but this may not always be the case. So, while we think, for this viz, labeler is a good proxy for “drugmaker,” we acknowledge that this won’t always be the case.
If you’re still upset about this, all we can say is we did our best with what’s available. The data on exactly where, and precisely who is manufacturing our drugs is flat out not available in the public domain (or even behind paywalls). Call the FDA and voice your displeasure… if you can get this serious problem fixed, we’ll gladly update our methodology. Again, as we’re doing the best we can with what’s available, please drop us a note if you see something that needs adjusted.
Stat Box #1: Cumulative WAC Price Changes (by week)
We’ve tried to pack a ton of information into the Brand Drug List Price Change Box Score. The idea was to include something for everyone… at least everyone who is into WAC pricing changes on brand name drugs and that sort of stuff.
Stat Box #1 is for the purists who simply are looking for the number of price changes each week, and each day. We simply count each WAC price increase as a 1 and each WAC price decrease as a -1, and then add everything up. The chart presents a running total of these “net” WAC price changes over the course of a year. Click on the “Select Year(s)” dropdown box to bring on more, or less, years for comparison purposes.
We’ve also incorporated a feature into this stat box to make it easy to find the daily price increases. Simply hover over any one of the series on the chart, and another visualization will pop up within the tooltip, showing the total number of WAC increases, WAC decreases, and median percent change on both the increases and decreases for each day in the selected week / year.
For example, the below screenshot shows the tooltip that pops up upon hovering over week 1 of 2020. From this box, we can then learn that there were 386 price increases on Jan 1, 2020, with a median price increase of 5.0%.
For journalists looking for daily updates on price changes (especially through January), this is the place you need to go within this dashboard to get that information.
FAQs
1) What is WAC?
WAC is the Wholesale Acquisition Cost, which is the “list price” brand drug manufacturers set for their products. WAC is not freely available but can be obtained through databases like GSDD, which 46brooklyn has licensed (Hint: Your donations enable us to purchase these types of licenses). We are permitted to publish changes in WAC, because we have developed algorithms to calculate the percent changes at the brand product level. We are not permitted to directly provide the specific WACs for drugs, which is why you will only find percentage changes in the visualization.
2) How are you defining a “drug” and how are you counting price changes?
We are using the following process and logic to define a brand name drug:
1) Use Elsevier Gold Standard Drug Database (GSDD) to find all brand NDCs (i.e. StatusName = ‘Brand’)
2) Filter the brand NDCs to only those with New Drug Application (“NDA”) or Biologics License Application (“BLA”) license types (i.e. LicenseTypeName = ‘BLA’ or ‘NDA’)
3) Remove all repackaged NDCs (i.e. Repackaged = ‘False’)
4) Remove all over the counter drugs (i.e. LegendStatusID = 2)
This produces a list of all NDCs that comprise the basis for the “brand” drug products presented in the visualization. We then concatenate the NDC’s trademarked name with the dosage form to create what we call a “brand product.” This is the level at which we have chosen to measure price changes.
We then aggregate all NDCs to the brand product level, and count price increases at this level. This means that all strengths and package sizes with the same trademarked name and dosage form will roll up.
To better illustrate, take Latuda as an example. As of the end of 2020, there are 11 NDCs on the market for Latuda, comprising five strengths and three different package sizes. On Jan 1, 2020, Sumitomo Dainippon increased the WAC of all 11 NDCs by 4.9%. Our process averaged these price increases across all 11 of the Latuda NDCs into one single 4.9% price increase on the brand product called Latuda Oral tablet.
3) What if there are multiple price increases in the same year?
So long that price increases have different effective dates, we will capture and report multiple price increases during the same year.
4) Why are you incorporating dosage form into your definition of a brand product?
Throughout our work at 46brooklyn, we have seen repeated instances of “product hops,” where a drugmaker will first bring one dosage form to market, and then “hop” to another one when the initial starts facing generic competition. Because of this pernicious dynamic, we felt it was critical to count different dosage forms as different drugs. We admit that this may be construed as “penalizing” drugmakers with several dosage forms for the same active ingredient. And that may not always be fair to do – there could be situations where such dosage form “swell” Is justified. Regardless, given the many problems with dosage form manipulation, we didn’t want users of the visualization to lose this level of detail.
Stat Box #2: Cumulative WAC Price Changes (by month)
Stat Box #1 is the granular way to look at price changes, giving you daily drill-down capabilities. Stat Box #2 is a more holistic view of price changes, showing you the total price increases by month over the last decade.
FAQs
1) What are the dividing lines in each year’s bar?
The dividing lines separate each month of the year. The months are organized chronologically from the left to right, i.e. January is the left-most slice of the bar, while December is the right-most slice.
2) Why are there percentage labels for January in each year?
When we put together this Stat Box, we found two things fascinating. First, of course is that the number of annual WAC price increases has declined rapidly since 2015. But the second is that the increases in January have remained stable, or even grown. What’s happening is that price increases are consolidating in January, considerably lifting the percentage of overall price changes that occur at the start of the year. We added the labels to January so you can see this immediately.
3) How can I tell what the total price increases are in each year? There’s no label on the chart.
Hover over any month to get more information not only on that month, but on the overall year. The chart below shows the tooltip that appears when we hover over January 2019. This tooltip tells you that there were 1,171 price increases in 2019, and that two-thirds of them occurred in January.
4) The math doesn’t seem to add up in the second sentence of the pop-up. What’s going on here?
Each day before we publish price changes we first run all of the data through a QC process to make sure units of measure (UoM) on each NDC are not changing. We noticed that - very rarely - this can happen in “hot off the press” updates from GSDD. Clearly, a change in a UoM can throw off the price increase, as it will erroneously change the unit price, which is what we are using to calculate price changes.
So if we suspect a UoM issue is causing a price increase, we identify these drugs and report them to GSDD. While GSDD is researching any issues, we exclude these drugs from all of the primary views across the dashboard. However, in the second sentence of Stat Box #2 pop-up box, we are using a complex level of detail calculation in Tableau to display the total annual increases. Apparently this level of detail calculation ignores the products we are temporarily excluding. As such, you may at times see a total increase count that is larger in Stat Box #2 than the one shown in Stat Box #1.
The good news is that this disconnect will get resolved over time as we work with GSDD on these potential UoM issues, and then bring the excluded drugs that are responsible for the disconnect back into the dashboard.
Stat Box #3: Cumulative WAC Price Changes (by month)
Hopefully Stat Box #3 is pretty self-explanatory, as it’s the simplest one in the box score. Stat Box #3 displays the median price change of all brand name drugs each year. Note that just as all the other stat boxes, it will display values only for one drugmaker when one is selected from the global filter at the top.
FAQs
1) Why are you showing the median and not the average?
Because it is misleading to take an average of an unweighted set of price changes. What if you have a 1,000,000% increase on a drug with zero volume in the U.S. market? That would massively skew the unweighted average to the upside but be of no consequence to our drug spending overall. As such, we see the median as a much more meaningful number for unweighted analysis.
Stat Box #4: Weighted Average Percentage WAC Increase
Stat Box #4 is, in our opinion, where this dashboard really starts to get interesting. Due to the difficulty in getting U.S. drug utilization data, really the entire discussion on drug pricing at the start of each year has been on the count and magnitude of price increases on individual brand name drugs. But think about this… would we use this method to try to understand price increases in any other marketplace? We could track all the individual price increases on items in, say, a grocery store, and take the median increase. But if people typically buy meat, cereal, eggs, ice cream, and blue milk, those price increases matter much more than say, the price increase on xanthan gum. Weighting the increases by what people buy and consume clearly makes much more sense.
This is no different with prescription drugs. The problem simply has been data availability. But the good news is that CMS does publish Medicaid drug utilization data in a granular enough fashion to pair up to price increases (in what they call their State Drug Utilization Database). So, we grabbed that, and merged it together with the price increases calculated based on pricing data in Elsevier’s Gold Standard Drug Database to tell a more complete story.
To take our grocery store analogy further, after weighting the price changes on all products in the store, what would we divide that number by? Would we divide it by the total spend just on those items with price changes? Or would we divide by the total spend in the entire store? The latter of course, as it doesn’t make sense to drop all of the products that didn’t experience a price increase. Yet historically when the general public talks about brand drug price increases, we only talk about the drugs that increased, and we don’t account for the ones that did not, again because its very difficult to create the data needed to have the right discussion.
To illustrate the impact of ignoring the zero-change drugs, we decided to display two lines in Stat Box #4. The blue line takes the total weighted value of all price changes, using prior year Medicaid utilization, and divides this by the total spending in Medicaid only on the pool of drugs that experienced a price change in the current year. For all brand name drugs, this resulted in a weighted average price increase of 5.2% in 2020. We then also display a yellow line, which has the same numerator as the blue line, but a dominator that accounts for spending on ALL brand name drugs in Medicaid in the prior year. As you’ll see, this drops the weighted average percent increase to 3.6% in 2020.
FAQs
1) How exactly are you calculating weighted average price increases?
We take the price change on each drug and multiply it by prior year Medicaid utilization on that drug. Say, you have a 5% increase in 2020 on $10,000 of 2019 utilization, that would be a $500 impact. We then do this for all drugs that experienced a price increase each year to get the total dollar impact to Medicaid. Note that if a drug has two price increases in one year, both are added to the numerator. Take the drug that had a 5% increase and $10,000 worth of utilization. If it had another increase later in the year of 3%, we would add another $300 to the impact to get a total of $800 for this one drug. Please note that this analysis does not consider the timing of the increase. Regardless if the increase is on January 1 or December 31, the math will output the same result.
Then, we create two different denominators. The first is the sum of all prior year spending just on drugs that had a price change. We take the numerator and divide by this number to get the blue line on the chart. We then create a second denominator that sums all Medicaid spending on brand name drugs in the prior year and apply this to the numerator to output the yellow line.
If you are a Tableau jock and are wondering how we display both on the same chart when they… well… we got very creative with calculated fields and level of detail (LOD) expressions. Feel free to contact us if you want more detail on exactly how we tricked Tableau into displaying both series on the same chart.
2) Doesn’t Medicaid get huge rebates on brand drugs? So why use Medicaid for this analysis?
Yes, Medicaid does get enormous rebates, making list prices essentially a moot point for the program that caps patient copays at no more than $8 for drugs. Ironically, due to the inflation penalty component of the Medicaid Drug Rebate Program (MDRP), Medicaid would actually prefer drugmakers to increase list prices faster, as that would result in lower net costs for the Medicaid programs. So, the waning in drugmaker price increases is likely in a roundabout way increasing Medicaid’s drug costs. And you thought big insurers, PBMs, pharmacies, wholesalers, and drugmakers were the only ones that like high drug prices?
But we digress. You are right, these list prices are rather useless in Medicaid. But Medicaid’s drug mix is NOT useless. Medicaid covers a large (and growing) chunk of this country, and therefore its utilization is quite useful from a weighting perspective. Yes, it’s not a perfect dataset for this exercise… there are many more children in Medicaid than your typical plan, skewing coverage to drugs for treatment of disease states like ADHD. Also, because of those massive rebates, it is common for Medicaid programs to cover a disproportionate amount of brand-name drugs relative to other payers in the market. This could “over-weight” some of the impact of some of these increases. On the flip side, CMS SDUD “suppresses” data on drugs with fewer than 11 prescriptions. This could likely mean that many of the pricey, niche biologics that have been hitting the market could be under-reported in the dataset. So yes, there are definitely some caveats in the utility of weighting price changes using Medicaid SDUD, but for the biggest two limitations, they at least generally counter-balance one another.
But regardless, Medicaid drug utilization data is quite robust, and therefore well worth using. Sadly, other Federal programs (Medicare, Indian Health Services, etc.) do not provide NDC-level reporting as an alternative for us to utilize in conjunction with Medicaid.
Plus, let’s consider the alternative. Not weighting price changes misses a lot of the brand drug pricing story. So now, we get to weight them and make our analysis a little bit more accurate. It’s like we’ve invented, built, and given you a car, when all you have ever known is a bicycle. We’ll admit, this car isn’t even close to a Mercedes… but it’s nonetheless a car and is way better than your bike when it comes to getting you from point A to B in your drug pricing research.
But this doesn’t mean we are satisfied with a ho-hum car. This country needs better than just relying on Medicaid’s drug mix to analyze weighted drug pricing. To that end, we implore CMS to create an NDC-level utilization database for Medicare Part D – and any and all other programs controlled by the federal government – that is free for the public to download (like SDUD). If CMS releases this data, we will gladly add other federal programs to Medicaid’s utilization to provide a better view into weighted inflation.
3) Why do you use prior year Medicaid utilization?
There are a few reasons we use prior year utilization. First, if Eric harkens back to his days running Financial Planning and Analysis (FP&A) for a Fortune 500 company, the correct way to measure pure price impact over a set period is to take the volume at the start of the period and multiply by that by the price change over the period. We’re not doing that exactly here (we are using the cost of the prior period), but the core concept still applies. As a brand name box score tracker, we believe the prior year reflects to a degree the manner with which the brand name labeler would be evaluating the impact its price increase may have in the coming year (they would have to rely upon prior year data when sizing the impact of a January 1 increase). Plus, it just felt more mathematically correct to use prior year utilization data to size the current year changes.
But we acknowledge this can be debated, and we could be convinced to change our minds on this topic. But what cannot be debated is that there is a substantial lag in reporting of Medicaid utilization. We only have it through Q2 2020 at time of writing, which is December 2020. So, the only way to weight 2021 price changes is with prior year utilization. We figured we may as well be consistent and do all years the same way.
By the way, if you were curious, to estimate full year 2020 drug utilization, we are just doubling the first half 2020 data we currently have.
Stat Box #5: Percent of Medicaid Brand Name Drug Spending Impacted by WAC Price Changes
After the painstaking process of creating the database to drive Stat Box #4, we needed to develop more views to look at the data in different ways. So, we created Stat Box #5 to complement Stat Box #4. Whereas Stat Box #4 answers the question of how different Medicaid list price inflation is on two different bases (only drugs with price changes vs. all brand name drugs), Stat Box #5 shows us the percentage of Medicaid spend each year on drugs that experienced price changes.
To calculate this, we simply took the two different denominators for the yellow and blue lines in Stat Box #4 and divided one by the other. Total Medicaid spending on drugs with price changes became the numerator, while total Medicaid spending on all brand name drugs became the denominator.
Overall, the chart clearly shows that the percentage of Medicaid spending that is “touched” by price increases is decreasing over time, falling from 87% in 2011 to 70% in 2020.
Stat Box #6: Medicaid Cost per Claim of Drugs with WAC Price Changes
Up until this point, we’ve really focused on two of the three “legs of the stool” – the number of price changes and the percent change. But there is a third leg of the stool, which is the starting price of the drugs being increased.
Of course, if the population of brand name drugs was a closed group, we wouldn’t need this leg… By “closed group” think of an exclusive club that you can’t get into (or out of). If the members in the brand name group were stable – like the members of this exclusive club – then we could just measure weighted price increases on these drugs and get a good representation of what’s going on with drug costs.
But the population of brand name drugs is not a closed group. It’s a fluid group where members are constantly exiting (i.e. when generics come to market) and entering (i.e. when new drugs are launched). With the group changing over time, there’s risk that we are not seeing the entire picture just looking at two of the three legs of the stool.
That’s why we created Stat Box #6 – to better understand what the drugs experiencing price changes cost in Medicaid. And it turns out that drugs experiencing price changes are getting considerably more expensive over time as new brand name drugs are launched at higher prices.
FAQs
1) How are these costs per claim calculated?
We first filter the database down to only those drugs that experienced a price change each year. For example, in 2019, there were 1,170 drugs that experienced a price change. We then sum up all the spending in Medicaid in the prior year (in this case, 2018) on these 1,170 drugs, and divide that by the total number of prescriptions dispensed on these claims. In 2019, this math all sorted out to be $601 per claim.
Stat Box #7: WAC Price Changes by Brand Product
Now let’s really get into the weeds. Up until Stat Box #7, we’ve just been playing with aggregates. Stat Box #7 gets granular, displaying prices changes on individual brand products.
Simply choose the week and year from the filters at the top, and the visualization will display all price changes throughout the week, sorted by date (chronologically) and drugmaker (alphabetically).
To help you sort through what drugs are more meaningful than others – at least from a cost perspective – we have added another filter called “Select Shading Value.” You can choose to color/shade the numbers either by Gross Medicaid spend on the drug, or by Gross Medicaid cost per claim. The darker the blue, the bigger the number.
For example, if you choose Week 1 2020, and set shading to gross Medicaid spend, you’ll immediately see a very deep blue highlight on the 7.4% increase in Humira Solution for injection. That’s because Medicaid spent a whopping $2.2 billion (before rebates) on this drug in 2019. Hover over the percent increase and you’ll find this number.
If you leave the week and year the same, but change the shading filter to gross Medicaid cost per claim, scroll down the list, and you’ll come across a drug called Revcovi, highlighted in very deep blue. Hover over this drug and you’ll see why – it cost $199,192 per claim in 2019 (again, before rebates)!
We added this shading function mainly to help you save time when you are trying to make sense of what matters and what doesn’t as you are shuffling through what in some weeks can be a very long list of price changes. But one of the side benefits of this function is it helps better explain the upward trend we see in Stat Box #6. What you’ll notice if you leave the shading filter set to gross Medicaid cost per claim and poke around enough through the weeks and years is that the deep blue drugs are getting much more expensive over time. This is due to escalating launch prices, a topic we wrote about towards the end of 2019. This is something for all drug pricing researchers and policy wonks to keep an eye on. It is now much more common for drugs to come to market costing tens or even hundreds of thousands per claim. An innocuous, say 3% price increase, ends up being not so innocuous when it’s applied to such a large number.
Stat Box #8: WAC PRICE Changes on Top Medicaid Brand Name Drugs
The last Stat Box #8 was really designed to separate the wheat from the chaff. Choose a year, and we display the top gross spending brand drugs in Medicaid from the prior year, from high to low. Right next to that, we’ll tell you the price increase, and its effective date. This is a much more direct way to see price increases on big-time spend drugs than simply by scrolling through and looking for dark blue shaded numbers in Stat Box #7.
FAQs
1) Why are there “Null” values in the date field?
A Null value in the date field means there was no price increase on the drug in the chosen year. These rows are important as they show you leading drugs that did not experience price increases.
2) What happens when there are multiple price increases on a drug in the same year?
When there are two price increases, we will display two lines of data with the date and percent increase for that drug. We also duplicate the number in the prior year Medicaid spend column. Take Symbicort for example, which experienced two price increases in 2020 – a 3% increase on 1/1/2020 and a 2% increase on 7/1/2020. As shown below, Medicaid spend will also show up twice. But to be clear, do not interpret this as $1.2 billion of spend. Symbicort spending was $610 million. This is just showing you that there was a 3% increase on the $610 million base, and then another 2% increase off the same base.
3) I get that, but why would one drug have both a Null field and a price change?
This is unfortunately an avoidable feature with how we put together this complex, multi-faceted database. As shown below, Januvia is the largest example of this “feature.” Januvia was increased by 4.9% on January 3, 2020, and there was $277 million in Medicaid spending in 2019. But what’s the deal with that Null row?
To answer this you have to first recall that we are starting all of our analysis at the NDC level, and then aggregating up to the Brand Product level. If you were to drill into Januvia’s NDCs, you would find that there was one sole NDC that did not have a price increase (thereby displaying Null). The reason why? It went inactive in the middle of 2019, so Merck didn’t bother to increase it in 2020. We drilled down to learn that the utilization on this one NDC was negligible in 2019… but it is still in the database, and therefore responsible for this Null row.
The confusing thing here is that since we have aggregated this database up to Brand Product level, we can’t show you the miniscule utilization on this one NDC or pull it out of the visualization altogether. So, the same overall 2019 dollars spent number shows up in that Null row, even though this is not truly correct.
Until we can figure out how to clean this little annoyance up, we’d urge you to just ignore those rows. From our research, they are driven by ultra-low utilization NDCs, that largely fall into this inactive camp. All that said, we would interpret the Januvia line as having one price increase of 4.9% on 1/3/2020 off a Medicaid spending base of $277 million.
4) Why can’t you show this for Medicare Part D?
Because CMS does not publish Medicare Part D drug utilization and spending data at the NDC level. Trust us, we want more than anything to not have to rely so heavily on Medicaid’s data. But until CMS makes broader data available, we must rely on Medicaid.
Questions
If you have any questions, comments, recommendations, or anything else, contact us here. Thanks!