Understanding the unit economics of genetic testing labs
Beyond handwaving explanations why labs are unprofitable — and what it means for the clinical genomics workforce
Welcome to Health & Wealth! If you’re new, you can subscribe for free to learn about genomics, life science companies, and my fresh insights on the innovation theme park of biology — delivered directly to your inbox:
Since my Invitae series part 1 & part 2, the landscape has significantly shifted over the past few months and I feel compelled to address the unit economics of genetic testing labs. My goal here is to provide fellow genetic counselors and ordering providers with a deeper understanding of the business of genetic testing. This is also for investors to understand in what ways we’ve been directionally correct but specifically wrong. As always, feedback is encouraged and welcome.
Happy reading! 🧬
Article Highlights
Recent hype in the sequencing market driven by product announcements from companies like Illumina ILMN 0.00%↑ and PacBio PACB 0.00%↑ overshadows the more sobering reality of layoff sprees and belt-tightening within genetic testing labs.
This has caused many on the frontlines of clinical genomics to question the financial sustainability of unprofitable commercial labs and its impact on patient care.
Reimbursement hell, capital-intensive tech stacks, race-to-the-bottom pricing, and broken economics of diagnostics development all make genetic testing a difficult line of business. It’s no wonder why many are struggling, especially in the current macroenvironment.
Most genetic testing labs are unprofitable today, but what ultimately matters is if a company can prove equity-efficient growth and a feasible road to future profitability.
Amidst the pain, it’s critical to recognize genetic counselors are the backbone of the genomic ecosystem for patients and providers. Medicare reimbursement and technology solutions are needed to deliver genetic services at scale.
Last month all eyes were on Illumina’s splashy announcement unveiling the release of the NovaSeq X series, a machine capable of sequencing the entire genome for $200. The Big Announcement Season for the sequencing market has continued at ASHG.
But behind the scenes in the genetic testing laboratory industry, a more sobering reality is acutely hitting the frontlines. Layoff sprees and belt-tightening are happening across the board — Invitae NVTA 0.00%↑ announced layoffs of over 1,000 employees (~one-third of their total workforce) and companies like Sema4 SMFR 0.00%↑ and Ambry Genetics have followed suit.
As the market increasingly lost patience towards profitless companies, many within the genetics community have started to question the financial sustainability of low-cost, high-volume clinical genetic testing — and where their own role fits in the current landscape.
A recent article addressing this moment of existential crisis postulating why genetic testing companies have been historically unprofitable has been widely shared in genetic counseling circles:
It’s true — genetic testing is inherently a difficult line of business that’s capital intensive, often undifferentiated and by and large a race to the bottom in terms of pricing. I 100% agree that insurance reimbursement is a major hurdle for commercial labs. The administrative burden and complexity of appealing and pushing for coverage add to the cost of genetic tests for everyone involved. From inconsistent reimbursement based on what coverage you have to outdated coding infrastructure, it’s a migraine-inducing mess.
But I also believe to simply conclude “the high cost for performing genetic testing necessitates high costs to patients and their insurance companies” is counterproductive. If we were all to put up our hands in the air and exclaim, “well, that’s just how it is, unfortunately,” we’d still be stuck in the era of $2,000+ tests looking at only one or two genes reserved for select patients.
To fully realize the potential value of clinical genomics for everyone, we need to make genetic information affordable and accessible. And that can only truly come from cost-effective genetic testing at scale. The question is whether companies can scale growth with strong unit economics.
So here I want to parse out the mission-critical forces that can make or break the long-term prospects of genetic testing companies. It behooves genetic counselors and ordering providers to understand the business of genetic testing on a deeper level because it impacts what information can be delivered to whom and ultimately, how we care for patients.
This piece is also for investors to understand in what ways we’ve been directionally correct but specifically wrong. Once we get over the veneer of hype and optimism, we can see where the true rate-limiting barriers to enabling precision medicine lie.
To understand the financial sustainability of commercial genetic testing labs, we need to understand the game at play:
Not all tests cost and are reimbursed the same
While $250 patient pay has emerged as the de facto out-of-pocket price for gene panels in recent years, we need to understand not all clinical genetic tests are created equal — both in how much it costs to perform testing and how much labs get paid.
How genetic testing labs get paid
In US healthcare, the user is typically not the buyer. Insurers make up the dominant proportion of buyers which can create misaligned incentives and non-transparent pricing — a whole separate issue beyond the scope of this post. But because of this, there’s naturally a fundamental tension between payors and labs — payors don’t want to pay more than they need to, and labs obviously want to get compensated for their services. Oh, and throw in clinician and patient-generated demand into the mix and you have one gnarly multi-dimensional Rubik’s Cube to solve for.
Especially for complex genetic tests, labs know that reimbursement policies and clinical guidelines can be crazy double-edge swords — they can limit or destroy business just as much as they can support and enable growth:
“A lack of or delay in clinical acceptance of broad-based panels such as our tests would negatively impact sales and market acceptance of our tests and limit our revenue growth and potential profitability. Genetic testing is expensive and many potential customers may be sensitive to pricing. In addition, potential customers may not adopt our tests if adequate reimbursement is not available, or if we are not able to maintain low prices relative to our competitors.”
[Invitae 10-K filing in 2019]
To add to the complexity, coverage and use of genomic testing are inconsistent across the country based on what insurance you have, and bizarrely, where the testing itself is done. It’s no coincidence that every lab tries to open a testing facility in North Carolina — that state has one of the most favorable coverage policies under its local Medicaid legislation.
I won’t pretend to know all the intricacies of insurance reimbursement (does anyone, really?) but here are two key anchors to understand:
Oncology is typically reimbursed better than other genomic specialties
High-margin tests can offset losses from low-margin tests for labs
Oncology is typically reimbursed better than other genomic specialties
Oncology almost always wins on revenue because it’s a high priority for payers (aka insurance companies). For example, Myriad breaks out volume and revenue by area. And oncology contributes more revenue despite having less than half of the volume of women’s health testing:
Furthermore, insurance reimbursement for somatic oncology tests such as minimal residual disease (MRD) or therapy selection tests is an order of magnitude higher than standard “run-of-the-mill” germline testing. For those unfamiliar, germline hereditary cancer testing identifies people with a higher risk of developing cancer (think BRCA1/2). Meanwhile, somatic therapy optimization or recurrence monitoring tests are specifically for patients diagnosed with cancer. Here’s a list of the 20 genetic tests with the highest medicare reimbursement rates in 2019:
Somatic oncology tests are high-margin tests because reimbursement rates can be much higher than what it costs to perform testing. Beyond momentum in scientific innovation, labs are also financially incentivized to push out new products in this space — quite simply because they can get paid at higher rates.
High-margin genetic tests can offset losses from low-margin tests for labs
Product diversification is a strategic game many genetic testing labs play. To survive, you can’t be a one-trick pony. While public companies don’t specifically break out gross margins for each of their tests, generally speaking, more established older products have higher margins than newly developed ones.
Ideally, labs should aggressively suppress the cost of goods sold (COGS) of older products to support the development costs of new products. These new products then hopefully expand the company’s revenue, ramping up to higher margins over time.
“There’s a constant balance between us driving revenue volume north and incurring the cost, particularly of new product introductions… Our mature product lines are pushing 70% gross margins. The new ones tend to be lower [gross margins], and even a product with good reimbursement.”
[Former Invitae CEO Sean George: Q3 2021 earnings call]
Why volume and vertical integration matters
Recognizing all the unglamorous realities of reimbursement hell, let’s go back to the original question I posed: can genetic testing labs actually scale growth with strong unit economics and eventually reach profitability?
While many factors are externally dependent, one important aspect a molecular diagnostics company can control is how much they invest in reducing COGS over time. And lowering costs in part comes from old-fashioned economies of scale and vertical integration.
Setting aside the business model itself, Invitae and other young labs undeniably pushed the envelope on exponential cost declines — from an average cost per test of $1150 in 2016 to $296 in 2021 all while driving up test volume.
Such significant cost declines have major ripple effects on the entire industry:
New entrants have it harder today — Nowadays, it’s unlikely you’ll see a new entrant with significantly lower COGS because they simply haven’t achieved testing at scale right out the gate. The bar is higher for young molecular diagnostics startups and success requires novel tech or science as opposed to sheer execution. In other words, you need to have something really special to even have a chance against these established core players (or interesting enough to get acquired by one of them).
Reimbursement will continue to fall as competition grows, leaving some to fall behind — The unspoken truth about insurance reimbursement for molecular diagnostics is that it’s like a limbo contest. Most insurers set an amount they pay for a test and most labs contract for the same rate. But insurers all expect price declines over time and will accordingly lower rates across the board. When that inevitably happens, labs that didn’t previously invest as heavily to lower COGS will see their unit economics crumble and find it difficult to compete on pricing. We’re already beginning to see this shake out, though I suspect we’re still in the early innings.
So the ability to lower prices as volume increases without relying so much on third-party vendors is absolutely necessary. But unfortunately for the daring, growth at all costs is difficult to sustain. In the past few months, many genetic testing companies have invariably realized the party is over and they need to cut low-margin tests to optimize for near-term profitability. Companies are now more selective with their bets instead of taking a “we offer everything” shotgun approach.
This brings us to my next point — being unprofitable is not within itself the problem. Here’s what I mean:
Is more revenue widening or shrinking margins?
A common misconception is that unprofitable automatically = bad. Respectfully, the analysis here in this article makes no sense:
“Although these companies are not generating operating profits, their investors aren’t necessarily hurting as a result. Stock prices for boutique, genetic testing labs don’t often sync with the lab’s financial health… For example, Invitae hit all time stock highs in December 2020 despite enormous losses reported in every quarter that year.”
First, many investors have been hurt plenty in this recent downturn. The diagnostics market has been absolute carnage over the past year, correcting itself from the irrational exuberance in early 2021. Stock prices reflect the market’s belief in a company’s future earning potential, not whether they’re unprofitable today. Most growth stocks were “in the red” for years before turning profitable — including companies you probably know or use every day: Amazon, Netflix, Tesla.
Second, yes, most genetic testing labs are unprofitable today. But what’s more important than simply looking at net income is gross margins and operating expenses over time. That is, as revenue increases, does the company have widening or shrinking margins? What you want to try avoiding is an unprofitable company with increasingly unattractive margins. It means the quality of revenue doesn’t improve as the company scales and instead incurs more losses as they grow.
So what do margins look like for publicly traded genetic testing labs? Let’s take a look:
If you’re more visual, this bar graph created by Maxx Chatsko gives you a relative sense of the latest (first half of 2022):
No singular financial metric tells you the full story about a company so this is absolutely Not Financial Advice™, but you can see commercial genetic testing labs fall into one of three camps:
Stable margins hovering around the 60-70% range
Improving margins steadily increasing over time
Deteriorating or unpredictable margins over time
The bottom line: Beyond being unprofitable, problems escalate for labs if margins get worse over time. The “go for broke” strategy becomes extremely precarious if a company can’t eventually prove equity-efficient growth at scale.
Why the economics of diagnostics development sucks
Okay, so commercial labs want to improve margins — part of that comes from scale and vertical integration, and another is product mix with a priority to develop high-margin tests.
But here’s the kicker about the economics of diagnostics development — it sucks.
Not only is it expensive to discover, validate, and distribute new types of assays, but also after you do all the heavy lifting, there’s limited IP exclusion to protect you against competitors. Unlike drug development, there’s little to no market exclusivity for clinical diagnostics and biomarkers.
People smarter than me have written about this, including Imran Haque, VP of Data Science at Recursion Pharmaceuticals (previously CSO at Freenome & VP of Scientific Affairs at Counsyl):
“In our capitalist system, these expensive trials will not be conducted if the private-sector funder does not think they’ll be able to make a positive return on the capital invested. The big challenge here is that a validation study isn’t meant to discover anything new; it just shows that something known works. In a vacuum, this makes these studies economically unfeasible: you’ll spend a bunch of money proving that drug / diagnostic X works, and then someone else can use your data to go off and sell X without having to invest in that study themselves (a “second-mover advantage” or “free rider problem”).”
What’s a lab to do about not being able to patent a biomarker itself? Many have turned to the “data moat” strategy that goes something like this:
build up a large proprietary clinical data set
patent the laboratory and algorithmic process of discovering biomarkers
validate the test as a “black box” to make it difficult for others to copy it
maintain a market advantage by improving the test with more proprietary data
Recently, the data moat strategy has been popular for multi-cancer early detection test development. But this also brings about its own challenges — how do you validate a black box? Imran goes on to propose ways to realign incentives for diagnostics validation — a piece well worth a read.
The bottom line: There’s a chicken-and-egg problem in diagnostics — labs rely on new products for revenue expansion, but biomarker discovery requires heavy R&D spend with limited IP protection. The economics of diagnostics development is broken and better incentive models are needed.
What this means for genetic counselors
I can’t write a catch-all resolution for all the problems I described above — that’s another lengthy discussion. But what I can address here is the question: what does this all mean for the genetic counseling profession? Labs cutting costs through clinical workforce layoffs certainly has a significant impact:
“A substantial portion of genetic counseling is now delivered through genetic testing laboratories who have packaged genetic testing with the offer of genetic counseling to draw in clients. If we see fewer companies maintaining genetic counselors on their staff, where will genetic counseling support come from for these patients?”
I agree — genetic counseling shouldn’t be viewed as an “add-on” subsidized by commercial labs to cover liability but as an integral part of supporting the genetic testing ecosystem for patients and providers. So let’s discuss solutions here.
Medicare needs to recognize genetic counselors as providers and reimburse them for their services
If you’re a Medicare or Medicaid patient and want to speak with a genetic counselor, Medicare (and some Medicaid programs) doesn’t reimburse such services because genetic counselors can’t bill as independent providers. Such unnecessary barriers lead to one too many stories like this:
Efforts are being made to enact the Access to Genetic Counselor Services Act — a long overdue step to supporting genetic counseling as a service, independent from genetic testing or physicians.
Technology can and should enable genetic counseling services to be delivered at scale
But we can’t just rely on linear strategies (ex. train/hire more genetic counselors) to solve nonlinear problems (ex. exponential number and complexity of genetic tests, new variants and validation research data, ever-changing guidelines).
To scale beyond our current capacity, we must reconsider how patients access genetic services and testing. Software and AI are well-suited to help frontline providers, especially given genetic counselors are a finite workforce.
No genetic counselor’s brain stores information on 150,000 tests, thousands of genes, and every detail about the current corpus of medical knowledge which doubles every 73 days. Nor should they be bogged down by tedious administrative tasks like printing and scanning genetic tests results into EPIC:
Not to sound like an overzealous techno-optimist, but hopefully it’s clear software solutions can help. I can expand in a separate post, but here are 3 promising areas to build solutions for:
Self-guided patient education and triaging — “Using technology in genetic counseling” has become synonymous with “chatbots” and “pre-test education videos” because it’s been brought up almost ad nauseam. But what we’re really asking is: does everyone need to see a genetic counselor before testing, or can technology make genetic testing more accessible in a medically-responsible way — without a 1-1 consultation? Examples: Gia chatbot, Genome Medical’s Genome Care Navigator
EHR-integrated clinical genomics decision support for providers — How can we make it easier for providers to create the most up-to-date care plans based on patients’ genetic results? The “last mile problem” is implementing genetics into clinical practice. We need ways to support clinical decision-making while saving providers’ time. Examples: Nest Genomics, Concert Genetics
Cohort-based genomics education for allied health workers — There are only ~5,000 genetic counselors in the US. That’s minuscule compared to other US healthcare workers (~1 million physicians, ~4 million nurses, ~700k medical assistants). Simple math: it’s much easier to upskill existing non-genetics healthcare providers than magically 10x the number of genetic counselors overnight. Novel edtech models could facilitate continuing medical education and genomics certifications for providers and medical trainees. Examples: Amboss (though I have yet to see one specifically for genomics education)
Lastly, let’s accept models that work in one setting might not work elsewhere with different resources and systems — solutions for 54gene in Africa are different than the UK with universal healthcare, which is also different than the US with… well, whatever it is we have as a healthcare system 😅
The big picture
Things are rarely ever black and white, and so too is why genetic testing labs are unprofitable. Genetic testing is a difficult line of business — from reimbursement hell to the required capital-intensive tech stack to the broken economics of diagnostics development. It’s no wonder why commercial labs have been struggling, especially during a market drawdown when the cost of capital is high and the macroenvironment has placed an accelerated interest in profitability.
But despite all the doom and gloom, I remain optimistic about the future of genetic counseling and genomic testing. You simply can’t enable precision medicine without first knowing your patient’s genetic information. What’s more, science is improving to measure things that tell you about disease states or risks we previously had no clue about. And that body of knowledge is exponentially expanding.
To be clear — what the industry is going through now is an important wake-up call to temper fairytale economics and focus on growth with strong unit economics. There are many minefields in this uphill battle against the status quo, but what’s the point of entrepreneurship if not to create value and solve problems?
Thank you to Carrie Haverty, Heather Kamen, Maxx Chatsko, Genomics Cow, and others for feedback on an earlier draft of this piece. My sincere gratitude to those willing to lend their time and expertise.
What am I wrong about or what might I have missed? I write to learn and sharpen my thinking, but also to invite others to correct my blind spots.
Enjoyed my deep dive on the unit economics of genetic testing labs?
Please let me know by hitting the ❤ button. It makes my day to see whether my readers like the content (it really does!)
Consider sharing this piece with your network, reach out with questions or feedback, and subscribe for more from me:
Until next time!
Christina
Well thought out and informative article. Thank you for taking the time to bring attention to these critical issues.
My only question would be about your comment around NC Medicaid and what you meant by it.
NC Medicaid is actually notoriously awful for genetic testing coverage since only CMA, a handful of oncology CPT codes (and not panels), and a smattering of a few other single genes are covered and reimbursable. No Tier 2 Molpath codes are covered/reimbursable, GSP codes, no WES, etc. However, their genetic testing coverage policy (the one you linked to) is open for public comment (again) until 01/23/2023 - many genetics professionals are and have been commenting and pushing for at least WES and a few panels to be covered (some day).
Maybe I’m just not understanding what you were trying to relay, and if so my apologies.