Generic drug prices take a tumble
Welcome to Fall
Fall has arrived here in our home state of Ohio. After a relatively warm and dry August and early-September, consecutive days of rain, wind and cold have dislodged amber, bronze, and burnt orange leaves in droves, driving them to the ground where they will lay to rest for the winter.
Much like these falling leaves, generic drug prices have dislodged from what, in hindsight, was a temporary resting place earlier in the year, and are now being driven into the ground by the unrelenting “storm” that is the generic marketplace.
We estimate that the price changes reported just this month translate to $244 million in annualized savings just on generic oral solid ingredient costs for Medicaid. That’s the largest monthly deflation number we have seen since we started tracking monthly generic drug price changes.
Let’s be clear about this. This doesn’t make any sense. This was not supposed to happen in the middle of a global pandemic. Prices were supposed to go up, exposing flaws in the highly interconnected global supply chain. And while there are more shortages than there used to be, and some drugs that have exposed the supply chain’s flaws, prices just never received the memo. Actually, the memo didn’t just get lost in the mail. Rather, it appears generic drug prices received an entirely different memo, instructing them to instead fall with the force of a 50-lb dumbbell (dumbbells, which by the way, did get the COVID-19-related supply chain issues memo as they now cost $115 each) tossed off a building.
What Happened?
On September 23, CMS published its latest update to the National Average Drug Acquisition Cost (NADAC) database. For the uninitiated, NADAC is the best national public database of surveyed pharmacy invoice costs to acquire prescription drugs. In other words, if you are looking to understand what pharmacies are paying to purchase drugs from their wholesalers each month (before rebates), and you don’t want to cough up a boat load of money to pay for it, NADAC is where you must go.
As we’ve discussed at length in recent months, NADAC is lagged by roughly two months. So the survey prices we just received at the end of September reflect pharmacy invoice costs from July.
We already mentioned that this was an epic month for generic deflation – so epic we didn’t initially believe our tools. But the great news is that each month we view this data through many lenses, interrogating it until we feel comfortable with the story it has to tell. Think of the rest of this report as a transcript of the interrogation of this dataset.
Anyway, if you’re into that sort of thing, read on.
1. This is the most deflationary price change histogram we have ever seen
Each month, we first look at how many generic drugs went up and down in the latest month’s survey of retail pharmacy acquisition costs, and compare that to the prior month (Figure 1).
Basically, the quick way to read Figure 1 is to look for blue bars that are taller than orange bars to the left of the dotted line, and exactly the opposite to the right of the dotted line. That would indicate a good month – more generic drugs going down in price compared to the prior month, and less drug prices going up.
Figure 1
Source: Data.Medicaid.gov, 46brooklyn Research
Calling this month “extremely good” is a huge understatement. First look at the bars to the left of the dotted line, starting with the 10-20% decrease category. Last month, 225 drugs decreased by 10-20%. This month that count rose to 398, a 77% increase. If you add up all of the drugs that decreased by more than 10% in both periods, there was exactly twice as many this month (blue bars) compared to last month (orange bars). That’s a huge swing … and remember, it’s not like last month was an “easy comp” (i.e. last month was not inflationary), this is just a reflection of the sort of beat-down that generic drug prices experienced in September.
Now let’s move to the right of the dotted line … a.k.a. the realm of increasing drug prices. NADAC price increases contracted by an impressive 25% from August to September. In September, for every drug that increased in price, there were 1.7 drugs that decreased in price. In August that ratio was 1:1.
The story gets even better though, as you’re about to find out in the next section.
2. Weighted Medicaid generic deflation tops the charts at $244 million
As we’ve written in prior updates, knowing the price changes alone are not enough. We need to apply utilization (drug mix) to the price changes, which is the purpose of the NADAC Change Packed Bubble Chart (Figure 2). We use Medicaid’s 2019 drug mix to arrive at an estimate of the total dollar impact of the latest NADAC pricing update. This helps quantify the real impact of those price changes from a payer’s perspective (in our case Medicaid, individual results may vary).
The green bubbles on the right of the Bubble Chart viz (screenshot below in Figure 2) are the generic drugs that experienced a price decline in the latest survey, while the yellow/orange/red bubbles on the left are those drugs that experienced a price increase. The size of each bubble represents the dollar impact of the drug on state Medicaid programs, based on utilization of the drugs in the most recent trailing 12-month period. Stated differently, we simply multiply the latest survey price changes by aggregate drug utilization in Medicaid over the past year, add up all the bubbles, and get the total inflation/deflation impact of the survey changes.
Figure 2
Source: Data.Medicaid.gov, 46brooklyn Research
This month is a wonderful example of why it’s so critical to weight drug pricing data. As compelling as Figure 1 was, it was just half the story. Figure 2 shows us the entire story. It shows us that not only did more drugs decrease than increase in price, but the drugs that decreased tended to be higher utilization drugs as well. Stated bluntly, teeny tiny bubbles means teeny tiny money (relatively) spent on drugs by Medicaid while large bubbles means Medicaid is spending lots of money on that particular drug. Similarly, green means savings and yellow/red means added costs to Medicaid. It just so happens that, with few exceptions, the “increase” side of the ledger is packed with teeny tiny bubbles, while Medicaid’s heavy hitter drugs (i.e. those with lots of utilization) are largely running with the “decrease” posse. Add it all up, and that explains how aggregate net deflation was able to eclipse $240 million this month.
3. Year-over-year generic oral solid deflation rockets up to 17%
Ever since last June, we have been tracking year-over-year generic deflation for all generic drugs that have a NADAC price. We once again weight all price changes using Medicaid’s utilization data. Over the past year, we have been seeing a gradual compression in deflation. But oddly enough, as the pandemic started, this trend started to reverse. This month, year-over-year (YoY) generic deflation on oral solids rocketed up to 16.6%, nearly one standard deviation above the trailing 24-month average. Meanwhile, deflation on all generics hit 12.2%, the highest level in over a year (Figure 3).
Figure 3
Source: Data.Medicaid.gov, 46brooklyn Research
4. The plummeting costs of ADHD treatment
The discussion on the pricing on ADHD drugs has been a fixture in our monthly NADAC reports this year, as month after month we have been mesmerized by their precipitous pricing declines. This month was no exception with steep drops on essentially all strengths of generic Concerta ER (methylphenidate ER) and generic Adderall XR (dextroamphetamine-amphetamine ER).
Figure 4
Source: Data.Medicaid.gov, 46brooklyn Research
At this stage, it’s worth taking a moment to size just how much Medicaid could save this year thanks to the free-fall in the acquisition cost of these two popular ADHD treatments.
To accomplish this, we pulled our handy dandy Elsevier Gold Standard Drug Database (GSDD) off the shelf (it rarely stays on the shelf for very long) and joined it together with Medicaid’s 2019 State Drug Utilization Database (SDUD). We used GSDD to filter down to the generic oral solids with generic names of “Methylphenidate Hydrochloride” (which is the Concerta family of drugs) or the mouthful of a drug name, “Amphetamine Aspartate, Amphetamine Sulfate, Dextroamphetamine Saccharate, Dextroamphetamine Sulfate” (the Adderall family of drugs). This returned a list of 205 National Drug Codes (NDCs), which rolled up to 41 different “product names”, where product name is the combination of the drug’s active ingredient, dosage form, and strength. Doesn’t sound like that many until you start tallying up the units, prescriptions, and spending in Medicaid for these two drug groups:
12,203,419 prescriptions
549,717,854 pills
602,971,044 dollars spent by Medicaid
And again, that’s just generics. This doesn’t even include brand-name Concerta and Adderall, which is also heavily dispensed in Medicaid thanks, in part, to rebates.
Our next step was to join in two NADACs – the last NADAC pricing sheet of 2019 (dated 12/15/2019) and the most current NADAC pricing sheet (dated 9/30/2020). We added both these pricing lists to the table and then repriced the total NADAC-based ingredient cost at these two points in time. Turns out that on 12/15/2019, the total NADAC of these drugs was $447 million. Yes, this is lower than the $603 million that Medicaid actually spent, but you have to remember that pharmacies theoretically are paid dispensing fees above their acquisition cost to cover their cost to operate a pharmacy. In this case, the $156 million markup works out to be $12.79 per prescription, not too far out of line with most state’s pharmacy cost to dispense surveys. So the markup we found here passes the sniff test.
Of course that’s all in theory. In reality, as 46brooklyn readers know quite well by now, managed care pharmacy benefit managers (PBMs) can take as much of that $12.79 per claim markup as they choose, whether directly through spread pricing, or after the fact through “generic effective rate” true ups. Without more transparency into how PBMs choose to divvy up the markups, it’s anyone’s guess how much is kept by the PBM and how much makes it to the pharmacy (further complicating matters, thanks to vertical integration, one company can be both the PBM and the pharmacy), which is why we can’t call it a “dispensing fee” and instead use the much more nebulous term, “markup.”
But while in the past, we have provided some detailed insights on how that markup can be split between PBMs and pharmacies, that is not the point of this little rabbit hole. The point is that when we repriced these same drugs with the same 2019 drug utilization using current NADAC prices, we got a total NADAC of just $277 million. Again, at the end of 2019, this same basket of drugs cost $447 million. This is the generic deflation storm we are talking about – it wiped $170 million in drug ingredient cost off of just two ADHD drug groups in just ten months. And if you are Medicaid, or just a taxpayer who is funding Medicaid, this $170 million should be yours.
If a state is running the show for its Medicaid program (i.e. “Fee for Service” or “FFS”), then it is guaranteed to see these savings, as FFS programs reimburse pharmacies at NADAC (or other comparable benchmark for the drug cost) plus some professional dispensing fee. In this model, when NADAC craters, so does the state’s cost. That’s not the case when the state farms out its Medicaid prescription expense to private insurers (i.e. “Managed Care Organizations” or “MCOs”), in which drug costs are typically driven by discounts to bogus and stale Average Wholesale Prices, rather than real acquisition costs. So, if any state Medicaid programs are reading this … if you fall into this second bucket, check your recent managed care ADHD drug claims receipts and make sure you see these savings. If they are nowhere to be found, call the managed care organizations to which you entrusted management of these costs, and ask them where the money went.
5. Keeping an eye on Generic Zoloft
Generic Zoloft (sertraline) is a medication used to treat major depressive disorder, obsessive-compulsive disorder, panic disorder, post-traumatic stress disorder, social anxiety disorder, and premenstrual dysphoric disorder. From the prior month’s NADAC to current, the price of the 50 mg tablet declined 32% to $0.05 per pill. This returns the pricing of sertraline 50 mg to its approximate pre-July 2020 pricing as shown in Figure 5.
Figure 5
Source: Data.Medicaid.gov, 46brooklyn Research
Back in May 2020, sertraline products went into shortage according to the FDA due in part to several manufacturers being unable to obtain ingredients to make the drug due to COVID-19. Unsurprisingly, prices spiked due to the shortage but now appear to be moderating (presumably as the shortage starts to resolve). Sertaline remains listed as a shortage on the FDA site, so it remains unclear whether this price change is durable.
6. ADPIT shows some warning signs for dexamethasone
Finally, let’s check in on our Abnormal Drug Pricing Increase Tracker (ADPIT) and see what it has to say about this month’s data. As a reminder, ADPIT takes all of the NADAC prices for one drug over any given 52-week period, ranks them, finds the 90th percentile for the price, and then compares the current price to that 90th percentile price. If the current price is above the 90th percentile, we consider the current week’s price to be “abnormal” and add it to the list for you to peruse. Then we do this for another 20,000 or so drugs for good measure to complete the list, shown in Figure 6.
If this is the first time you are seeing this, we include the number of annual Medicaid prescriptions for each of these abnormally priced drugs, and then two metrics we created – one called Relative Impact Ratio (RIR) and another called Relative Impact Score (RIS). RIR simply tells you how far above the 90th percentile a drug’s price currently is (i.e. Dexamethasone 4 MG/ML Vial at 1.38 means its 38% above its T-52 week 90th percentile price). RIS multiplies the number of annual Medicaid prescriptions by (RIR - 1) to size the impact of the “pricing abnormality” (for dexamethasone, 38% x 325,516 prescriptions = 124,072). Add up all the RIS for all of the drugs each week, take the four-week moving average, and you get the chart in bottom right of Figure 6 (which sizes the aggregate weighted impact each week). We then divide the bottom right chart into drugs with active ingredients in ongoing COVID-19 clinical trials, and all other drugs. Check out our interactive ADPIT tool here, and find more information on how to use it here and here.
Figure 6
Source: Data.Medicaid.gov, ClinicalTrials.gov, 46brooklyn Research
Overall, the dashboard paints a very healthy picture for the generic marketplace. The chart on the bottom right shows the aggregate RIS is at historically low levels, and the leaderboard doesn’t highlight any high utilization drugs that have blown through their T-52 week 90th percentile price.
But there is one very noteworthy and concerning finding, and that is this month’s price spike in COVID-19 treatment dexamethasone. Dexamethasone is a steroid that reduces inflammation in several medical conditions, including those that affect the lungs. As shown in the chart in the bottom left of Figure 6, dexamethasone vials spiked 137%, from $0.59 per unit to $1.39 per unit. That’s a serious jump for sure, but take a look at the historical volatility of dexamethasone. Sharp increases and retractions in its price are not uncommon.
Owing to this extreme historical pricing volatility, it may be too early to sound warning bells on dexamethasone prices. Nonetheless, this is still a concerning finding in light of, 1) its growing importance in treating COVID-19, 2) that it is reported in shortage by both the FDA and ASHP, and 3) the fact that President Trump recently received treatment with dexamethasone as part of his ongoing COVID-19 diagnosis. Recall that when President Trump was previously promoting hydroxychloroquine for COVID-19, we saw a massive surge in generated prescriptions for hydroxychloroquine which were followed by price increases. Given the impact previously seen with other COVID-19 drugs associated with President Trump, this dexamethasone situation is something that federal officials will need to keep a very close eye on going forward if for no other reason than the fact that CMS data also suggests that the current utilization is heavily concentrated in a few drug labelers.
As seen in Figure 7 from our Medicaid Drug Market Share Dashboard, which tracks drugmaker market share using CMS drug utilization data (most recently from Q3 2019), it demonstrates that 70% of all injectable dexamethasone sodium phosphate solution utilization (i.e. the group of dexamethasone 4mg/mL vials belong to) is concentrated in just one labeler – Fresenius Kabi (the orange slice of the pie in the upper right of the chart).
Figure 7
Source: Data.Medicaid.gov, 46brooklyn Research
Given the historical correlation between COVID-19 treatment hype and increased treatment prescription generation (and the likelihood of follow-on price increases as supply and demand are pressured), there is good cause for pessimism on the future of the accessibility and pricing stability of dexamethasone.
As a comparison point, before the pandemic hit, hydroxychloroquine prices had been relatively stable. As we showed back in May 2020, as the surge on hydroxychloroquine hit, and doctors began issuing highly questionable prescriptions for dogs and arthritic goldfish, we saw NADAC prices jump 32%. However, the spike was short-lived, as we highlighted in our July 2020 NADAC report. Bear in mind that while hydroxychloroquine weathered the storm, its value as a legitimate treatment for COVID-19 was highly speculative and eventually fell flat. Here, we see a drug that rather than being touted by President Trump as a possible treatment, is a drug that is being used as an actual treatment. The perception of dexamethasone (and other tandem therapies) as a miracle drug, now that the President has quickly emerged from the hospital where he was originally admitted, throws added intrigue for anyone speculating on the possible added pressure that can hit the dexamethasone supply as a result of its heightened profile. Add to the fact that we are already seeing price spikes of dexamethasone two months prior to the President being prescribed the drug, as well as the fact that the supply of the drug is coming from a manufacturer marketplace with some relative concentration risk, we could have a recipe for disaster for the future accessibility and affordability of this crucial treatment if the supply chain does not hold up the same way hydroxychloroquine did.
Thanks to John Wilkerson at Inside Health Policy for his write-up on our Australian drug pricing report and our analysis of the current non-nonsensical push for tying US drug prices to an international price index. Additional thanks to Rhiannon Collette at Arnold Ventures for her excellent overview of the nuances of that same report. In case you missed it, you can read that report here.
Shout-out to Andrew Siddons at Roll Call for discussing the intricacies of drug pricing and the supply chain with 46brooklyn’s Antonio Ciaccia in his analysis of President Trump’s record on tackling the nation’s rising prescription drug costs.
Lastly, make sure you check out Noam Levey’s eye-opening piece in the Los Angeles Times, where in joint a analysis with 46brooklyn, he looked into the irregular midyear drug price increases from AstraZeneca, which occurred parallel to the drugmaker’s hundreds of millions of dollars in taxpayer subsidies for COVID-19 vaccine development. It was an honor to work with Noam and the team at LAT on the collaborative data analysis.