This episode reviews a claims-based study of 241,060 Medicaid-enrolled children (ages 7–17) from 2008–2016 that measured use of genetic testing among those with ASD-only, ID-only, and ASD+ID. The authors report low overall testing rates, temporal shifts in test modalities from cytogenetics/Fragile X toward chromosomal microarray and gene panels, and disparities by race and urbanicity. We outline the methods, key findings, and clinical implications for guideline implementation and access to genetic services.
0:00Welcome to Base by Base, the papercast that brings genomics to you wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. Thanks for having me back. So I want you to imagine, holding the key to understanding a child's complex neurodevelopmental condition.
0:16Right. And it isn't just like a vague label. It's a physical key, they could unlock highly targeted treatments, specialized medical care, or even down the line, future gene therapies. Yeah, a real game changer for a family.
0:30Exactly. But now, imagine that for the vast majority of children relying on the public safety net, that key is just left entirely unturned. Which is incredibly frustrating. It is. Because the craziest part of this whole scenario is that the key actually works.
0:45Like genetic testing for intellectual disability in autism spectrum disorder has a massive diagnostic yield of, what, 25 to 38%? Yeah, that's right. Between 25 and 38%. But despite that incredible success rate, the children who need it most simply aren't getting it.
1:01So what really happens when clinical guidelines outpace real world healthcare access? And how could this massive gap change the entire future of precision medicine? Well, today we celebrate the work of a multi-institutional research team, including researchers from UCLA, Drexel University, the National Institute of Mental Health, and Mount Sinai, who have advanced our understanding of the clinical implementation of genetic testing among underserved populations.
1:25And to really grasp the gravity of this data in our deep dive today, we need to, um, kind of understand the human biology at play first. autism spectrum disorder and intellectual disability effect roughly one to 3% of the US population.
1:40And historically, diagnosing the underlying cause of these conditions felt like, I don't know, throwing darts in the dark. Oh, absolutely. It was largely based on observing behavior and physical traits.
1:52Right. And I know genetics has completely changed that landscape, but how exactly are we getting to that 25 to 38% diagnostic yield you see in the current literature? Well, it really comes down to a massive leap in technological resolution.
2:05Because for decades, we knew these neurodevelopmental conditions had a biological basis, right? We just, we couldn't see the specific errors in the code. didn't have the tools. Exactly. But today, we know that single nucleotide variants, which are, um, tiny microscopic typos in a single DNA building block and copy number variants, where whole paragraphs of the genome are accidentally duplicated or deleted, they play a dominant role.
2:34Wow, whole paragraph. Yeah, and finding those specific variants is what gives doctors the ability to definitively diagnose the root cause of the condition. I remember reading that the medical establishment actually recognized this shift relatively quickly, right?
2:46They did, yeah. Like from 2000 to 2007, the standard recommendation from pediatric academies was mostly just karyotyping and fragile X testing. Right, the older methods. And that was reserved for highly specific cases.
3:00Usually involving distinct physical features or a clear family history. It wasn't for everyone. No, not at all. And we should probably clarify the mechanics of standard karyotyping, just so you understand how limited it was.
3:10please do. It is essentially 1950s technology. Kareotyping is like standing across the room and looking at a massive multi-volume encyclopedia set on a shelf. Okay, so you can see the big picture. Right.
3:23You can tell if an entire book is missing or if an extra volume was crammed in there, but you cannot read the pages. can't see the typos. Exactly. You certainly can't tell if a specific paragraph was deleted.
3:33Which totally explains the paradigm shift in 2010. The American college of medical genetics. or ACMG, publish this massive update to their practice guidelines. They explicitly recommended chromosomal microarray analysis, or CMA, as a 1st tier test for unexplained autism and intellectual disability.
3:52Right, CMA became the new standard. Suddenly, we had a tool that acted like a magnifying glass, scanning the actual pages of the genome for those microscopic copy number variants. And then by 2021, the guidelines evolved again, pushing for even higher resolution scans like whole XOM and whole genome sequencing.
4:09Yeah, the technology just kept accelerating. But, and this is crucial. This brings us to the core tension of this entire analysis. The gap. Exactly. Because having a progressive medical guideline published in an academic journal is entirely different from a child actually receiving that scan in a community clinic.
4:27Oh totally. Previous studies attempted to measure this adoption rate, but they were, frankly, deeply flawed. How so? Well, they relied on tiny clinical samples, or they asked parents to recall the specific names of complex medical tests their children received like 5 years prior.
4:44Oh wow. I couldn't tell you what exact blood test I had 5 years ago. Right. It's not reliable data Wait, I also saw that most of those older studies predominantly looked at privately insured populations.
4:56Doesn't that completely skew the data? Oh absolutely. I mean, why would researchers intentionally ignore a massive chunk of the population who rely on public health systems? It's usually a matter of convenience or resource availability, but that blind spot is exactly why this new study is so vital.
5:12Right. If we look at the adoption of a medical breakthrough, but only survey people living in affluent neighborhoods with premium insurance. We aren't really measuring the healthcare system. No, you're measuring privilege.
5:24It's like looking at the adoption of electric vehicles, but only serving people in high income neighborhoods. We need to know what's happening everywhere else. That's a perfect analogy. And Medicaid is the public health insurance program for people with low incomes.
5:37It serves a population with immense racial and ethnic diversity. So it's a completely different demographic. Exactly. These are the demographic groups most vulnerable to systemic healthcare barriers. If chroma zomal microarrays are only reaching the privately insured, then precision medicine isn't truly integrated into our society.
5:57That makes a lot of sense. So to find out what was actually happening within that safety net, the researchers bypassed clinical surveys and went straight to the billing data. Right, the hard numbers. They use the transformed Medicaid statistical information system, or TMSIS, they tracked longitudinal claims data from January 1st, 2008 all the way to December 31st, 2016.
6:17And they built a cohort of 241,060 children, all aged 7 to 17 in the year 2016. But they applied an incredibly strict filter. Very strict. These kids had to be enrolled in Medicaid for at least 9 months out of every single year, and they could not have a single day of private insurance.
6:35Wow, not a single day. Why be that restrictive? They needed to isolate the pure Medicaid ecosystem. They wanted to ensure the data wasn't contaminated by private healthcare resources. Got it. So the kid had private insurance for a month, they were out of the study.
6:50Exactly. Once they isolated that massive group of a quarter 1000000 children, they sorted them into 4 distinct categories based on their diagnostic billing codes. Okay, what were the groups? An ASD only group, an ID only group, a group with both autism and intellectual disability, and finally, a baseline random sample of kids without either a diagnosis.
7:10Okay, and then they scoured the records for current procedural terminology or CPT code to see who actually received a genetic test. Right. The codes for cytogenetics, micro arrays, and gene panels. But um, I am genuinely stuck on the timeline here.
7:24The data window abruptly stops in 2016. Why look at a window that ended almost a decade ago? Why are we analyzing this to understand modern genomics? I understand the confusion, but the timing is actually the most critical feature of the study design.
7:40Really? How so? Well, think back to that massive ACMG clinical guideline shift we just discussed, the one published in 2010. Oh, when they told every doctor in the country to start using micro arrays as a 1st tier test.
7:53Precisely. This 2008, 2016 window operates as a perfect real world laboratory. Okay, I think I see where you're going with this. It captures the baseline years just before the guideline was published, the immediate shockwave of the publication, and then the long adoption period afterward.
8:10Oh, wow. So it forces us to measure exactly how long it takes for a mandate issued by a national medical board to actually change the behavior of a primary care doctor typing into billing code. Exactly.
8:22It measures the lag. And if the goal was to measure clinical adaptation. The findings are incredibly sobering. Let's look at the actual testing frequencies. quite low. Yeah, you have national guidelines explicitly telling doctors to order these tests.
8:35Yet the group that received the most testing, the children diagnosed with both autism and intellectual disability, saw a testing rate of only 25.94%. Apparently, one in 4 children with the most complex clinical needs received the standard of care.
8:48And that was the best case scenario. The autism only group sat at 16.86%. The intellectual disability only group was at 13.09%. And the random sample baseline was just one. 23%. I mean, for diagnostic tool that is supposed to be the 1st tier response, nearly 80 to 90% of these children are completely falling through the cracks of the medical system.
9:12It's alarming, but to contextualize those low frequencies, the researchers calculated odds ratios. Okay, break that down for me. Well, children with both ASD and ID had 29.43 times higher odds being tested compared to the random sample.
9:26Okay, so that's a huge jump. It is, and this tells us something crucial about physician behavior. The healthcare system is attempting to prioritize. Doctors recognize that children presenting with the highest burden of clinical impairment desperately need biological answers.
9:40I see the logic there, but prioritization feels like a hollow victory, if the total pool of testing is just a trickle. Prioritizing who gets a life raft. Doesn't change the fact that the ship is sinking and there are only a few rafts.
9:54That's a very stark way to put it, but it's accurate. And when we look at the demographic breakdown, the allocation of those resources reveals stark, uncomfortable disparities. Let's examine the racial data.
10:06Yeah, the claims data shows that black children had significantly lower odds of receiving a genetic test across all cohorts when compared to white children. Wow, across all cohorts. Yes. For example, in the autism only group, black children face 24% lower odds of being tested.
10:22That is substantial. It is, but curiously, the data also revealed that Hispanic children had consistently higher odds of being tested across all cohorts compared to white children. Really? I know this claims data can explicitly tell us why those racial disparities exist, but it strongly implies that healthcare access is fractured along cultural and systemic lines.
10:41Absolutely. The structural barriers affect different communities in very different ways. And we see a similar fracture when we look at the geographic data, right? Yes. Children living in suburban and rural zip codes had significantly lower odds of being tested than children in urban centers.
10:58Which makes sense if you consider the physical reality of rural healthcare. Urban areas house massive academic medical centers with dedicated pediatric genetics clinics. Exactly. If you are a Medicaid enrolled family living in a rural area, obtaining a micro array isn't just about getting a doctor's permission.
11:16It requires finding a rural primary care provider who is familiar enough with the 2010 guidelines to even initiate the process. And then that family has to secure transportation, potentially take multiple unpaid days off work and drive hours to an urban specialist, only to navigate a labyrinth of new referrals.
11:34The friction is completely geographic. The logistics alone are a nightmare. Speaking of friction, The study found a highly nuanced disparity regarding sex, didn't it? Yes, this was a really interesting finding.
11:46Females with an autism only diagnosis, had lower odds of getting tested than males. Okay. But females with an intellectual disability only diagnosis actually had higher odds of being tested than males.
11:58That feels contradictory. Why would it flip like that? Well, this specific disparity likely stems from how different conditions present physically and behaviorally, the phenotype. Oh, I see. Because autism has historically been diagnosed more frequently in boys, a female presenting with autism symptoms might face clinical skepticism.
12:17So that delays her referral for genetic testing. Exactly. Conversely, certain genetic causes of severe intellectual disability in females might present with very distinct physical characteristics. Like visible symptoms.
12:30Right. And those distinct traits might immediately trigger a physician to order a genetic workup. So the visible symptoms dictate the doctor's response. But honestly, what blew my mind wasn't just who the doctors were testing, it was what they were testing them with.
12:44Oh, yes. The technology lag. Yeah. This is like doctors prescribing a pager years after the smartphone was invented. The ACMG told everyone to use chromosomal microarrays starting in 2010. But the billing data shows that all the way through 2013, the vast majority of tests built to Medicaid were the older, low resolution cytogenetics, the karyotyping, and fragile X tests.
13:11Right, the 1950s technology. It wasn't until 2014 and 2015 that the newer micro rays finally became the dominant test. Why would doctors spend years ordering outdated technology? You really have to look at the administrative mechanics of a clinic.
13:25This is a classic example of clinical inertia. Clinical inertia. Yeah. When a doctor sits down at a computer to order a test, they're using an electronic health record system. Often, the CPT billing codes for the older carriotype tests are just saved in their default dropdown menus.
13:40Oh, wow. So it's literally a software habit. Partly. Yes. Ordering the new microarray requires learning a new code, justifying it to the insurance provider, and potentially facing a rejected claim if they fill out the prior authorization incorrectly.
13:52So they just click the old code because it's what they know and it's what the system easily accepts. Exactly. It's the path of least resistance for an overworked doctor. But if doctors are stuck in a 2007 mindset for 5 years, that creates a terrifying math problem for a child in the waiting room.
14:08It does. I mean, a 13 year old waiting for the right test doesn't just stay 13. By the time that local clinic finally updates its software and orders a micro array, that kid is aged. And the data reflects that tragedy perfectly.
14:22Less than 6% of the first instances of genetic testing happen when these kids were between 15 and 17 years old. Less than 6%. That's incredibly low. Because neurodevelopmental conditions require early intervention.
14:35If a child misses the testing window in their early pediatric years due to clinical inertia, geographic distance, or a doctor using an outdated dropdown menu. They are highly likely to age out of the pediatric system entirely without ever receiving a biological diagnosis.
14:51And once they age out. Once they enter the adult medical system, specialized genetic screening becomes drastically more difficult to obtain. I'm trying to synthesize all these systemic failures. We have space age genomic technology capable of finding microscopic DNA typos.
15:06We have national clinical guidelines demanding doctors use it, but in the Medicaid system, the implementation is just fundamentally broken. We are observing a severe translation gap. The core issue is the profound administrative friction required to move an evolving medical recommendation into an equitable on the ground clinical practice.
15:27Let's break down that friction. Because it's not like community pediatricians are actively trying to withhold care from their patients. No, not at all. The barriers are baked into the bureaucracy of Medicaid itself.
15:37It's not a unified system with a blank check. Every single U.S. state establishes strict labyrinthine criteria that a patient must meet before a micro array is covered. Right. A primary care doctor can't just swab a cheek.
15:51They have to secure specialized referrals. They have to fill out extensive multi-page prior authorization forms just to legally justify the medical necessity of the test to the state. It's like having a hyperadvanced diagnostic machine, sitting in a lab, but forcing the patient's family to navigate a 15 page fax machine process to unlock the door.
16:11That's exactly what it's like And for a low income family, juggling multiple Doms, managing a child with a severe disability, and potentially dealing with language barriers. Navigating that administrative maze alone is nearly impossible.
16:23They require service brokers. Right. Right. Care managers or social workers to literally guide them through the paperwork. If your clinic doesn't have a social worker, you just don't get the genetics test.
16:33And we really must emphasize why failing to secure that test is so devastating on a clinical level. A genetic diagnosis is not simply a label meant to satisfy a parent's curiosity. It is the absolute foundation of precision medicine.
16:47Right, because knowing the exact copy number variant changes the daily reality of the patient, if the micro ID flags a specific genetic deletion, the doctor might instantly realize that this specific child is at a massive risk for sudden seizures.
17:01Exactly, or that they require an echocardiogram every 6 months to monitor a known heart defect associated with that exact gene. It literally dictates the entire surveillance regimen. Furthermore, it opens the gateway to clinical trials and syndrome specific interventions.
17:17We are entering an era of targeted gene therapies. If a child does not have their specific genetic variant documented in their medical file, they are permanently locked out of those future cures. Okay, so a test gets missed, and I understand the tragedy for a single family, but the paper suggests this failure is actually threatening the future of science itself on a macro level.
17:39It really is. I'm struggling to see how a missed test in a local Medicaid clinic derails global genomic research. Well, think about where scientific research originates. It relies on massive aggregated databases of genetic information to understand disease mechanisms and develop new drugs.
17:55Okay, the big dat banks. Right. If low income and minority populations are systematically blocked from receiving genetic tests today, because of administrative friction, their DNA is never entered into those research databases.
18:07Oh, I see the cascading effect. If the databases are only populated by the genetic information of privately insured, upper middle class individuals who could afford to bypass the friction, then our entire understanding of human genomics becomes skewed.
18:21Wow. We end up developing drugs and therapies, optimized for a very specific demographic, fundamentally limiting the generalizability of every scientific breakthrough we achieve. Precisely. We are literally building the future of precision medicine on an exclusive non-diverse foundation.
18:39That is terrifying. But to maintain scientific rigor, we should briefly touch on the study's limits, right? Yes, we must acknowledge the inherent limitations of the methodology that brought us to the conclusion.
18:50Claims data is incredibly powerful for observing systemic trends, but it has distinct blind spots. It does. It shows us what a doctor successfully build for. It cannot tell us the clinical reasoning behind why a doctor chose not to order a test.
19:04Right. It also can't capture tests that were billed under miscellaneous or unspecified codes. And because the researchers intentionally excluded anyone with private insurance to get a pure look at the safety net.
19:15We don't have a direct in study control group of affluent children to definitively measure the exact width of the access gap. We just know the Medicaid numbers are staggeringly low on their own. So taking all the CPT codes, the guideline lags, and the demographic disparities into account.
19:32How do we distill this massive data set? If we synthesize the findings, The core inside is undeniable. Despite clear national clinical guidelines, an alarmingly low percentage of Medicaid enrolled children with neurodevelopmental disorders are actually receiving recommended genetic tests.
19:48Stark racial and geographic disparities dictate who gets access, proving that our current system fails to translate advanced research into equitable clinical practice. Consequently, our most vulnerable populations are being systematically left behind in the precision medicine revolution.
20:04It is a profound reality check on the limits of our medical advancements. We have mapped the human genome. We have the technology to read it, but we have built an administrative maze around the laboratory that only a privileged few can navigate.
20:19It's a sobering truth. Which leads us with a final lingering question for you to think about. What does this mean for the future of human genomics if the foundation of our data only represents those privileged enough to overcome systemic healthcare barriers?
20:33This episode was based on an open access article under the CCBY 4.0 license. You can find a direct link to the paper and the license in our episode description. If you enjoyed this, follow or subscribe in your podcast app and leave a 5 star rating.
20:48If you'd like to support our work, use the donation link in the description. Now stay with us for an original track created, especially for this episode, and inspired by the article you've just heard about.
20:57Thanks for listening, and join us next time as we explore more science, base by base.
