Here is one of the great achievements of twentieth century medicine, stated as a problem.
Children who used to die now live. Congenital heart disease, cystic fibrosis, sickle cell disease, spina bifida, childhood cancer, complex metabolic disorders. Conditions that were once uniformly fatal in childhood are now chronic conditions of adulthood.
The children grew up. The expertise did not follow them.
A 24-year-old with a Fontan circulation walks into an emergency department with abdominal pain. The physician who sees her is competent, well trained, and has almost certainly never managed a Fontan patient. Her cardiology is not a variant of normal cardiology; it is a different physiology with different failure modes and different rules about fluid, rhythm, and anticoagulation.
Somewhere there is a cardiologist who has managed hundreds of patients exactly like her. There are, in the entire United States, about five hundred such physicians.
The emergency physician has no way to find one, and no reason to know they are a distinct category.
The arithmetic of adult congenital heart disease
This field has the cleanest numbers, so it makes the best illustration.
- There are approximately 1.5 million US adults living with congenital heart disease, growing at roughly 5 percent a year.
- There are 509 ABIM-certified adult congenital heart disease cardiologists.
- That is roughly 500 patients per board-certified physician, and about 3,300 patients per provider counting the broader care workforce.
- Fewer than half of boarded ACHD physicians practise ACHD full time.
- Fellowship output is insufficient to offset attrition, and projections suggest that even if training doubled, the ratio would still be around 1,600 patients per cardiologist by 2050.
Now the consequences, which are measured and specific:
- 42 percent of 922 ACHD patients had a gap of more than three years in specialist care, with a common reason being simply not knowing follow-up was required.
- In a 2004 to 2018 cohort of 2,196 referred patients, 25 percent were lost to follow-up and 11 percent failed to transfer at all.
These are not patients who refused care. They are patients who fell out of a system that never had a mechanism to hold them.
The pattern repeats across every pediatric-onset condition
Sickle cell disease. The infrastructure has contracted in a way that is genuinely startling. Single-specialty adult hematology programs fell from 74 in 1995 to two in 2018. Classical hematology, the non-oncology side of the field, has been substantially absorbed by oncology, and 46 percent of hematologists report a shortage in classical hematology.
The result for patients: in one California analysis, 49.1 percent of adult Medicaid patients with sickle cell disease and chronic kidney disease had no hematologist visit over a decade. Separate work found 46 percent in California and 21 percent in Georgia saw neither a hematologist nor an experienced primary care physician.
A patient in qualitative research put the situation plainly: "some hematologists don't even deal with sickle cell." Another summary from the same literature: "Not enough sickle cell clinics. Not enough time. Not enough providers."
Meanwhile, curative gene therapies have arrived for sickle cell disease, delivered at specialized centers at extraordinary cost. The therapeutic breakthrough has raised the price of not being able to find an expert.
Childhood cancer survivorship. Adult survivors of childhood cancer carry specific long-term risks, particularly cardiac, requiring surveillance that is well described in guidelines. In one vignette-based study, general internists failed to recommend guideline-concordant cardiac surveillance in 85.1 percent of cases.
That is not a knowledge failure by careless physicians. It is a structural one: the guidelines exist, the internist has no reason to know this patient belongs to a surveillance category, and nothing in the system flags it.
The deeper problem: the map has the wrong axes
Now the structural diagnosis, which generalizes well beyond these conditions.
Medical certification and every directory built on it partition the profession by organ and by age. Cardiology. Hematology. Pediatrics. Internal medicine.
Pediatric-onset conditions in adults cut across both axes. Adult congenital heart disease is not pediatric cardiology and it is not adult cardiology; it is a distinct expertise that happens to have earned a certification, which is why we can count its 509 practitioners at all.
Most conditions in this category have no such recognition. There is no board for adult sickle cell care. The internist managing an adult with a metabolic disorder diagnosed at eighteen months is working outside every category the system knows how to name.
Expertise organized by organ and age cannot describe expertise organized by condition and life stage. And what a taxonomy cannot describe, a directory cannot index, and a referring physician cannot find.
The same failure, one category wider: hyphen specialties
Once you see the shape, you find it everywhere in modern medicine, and increasingly in its fastest-growing areas.
Consider the patient who is pregnant with lupus and a mechanical valve. The cancer patient with trastuzumab cardiotoxicity. The transgender patient needing hormone management alongside a complex comorbidity. The 28-year-old needing fertility preservation before chemotherapy begins next week.
Each requires expertise that sits between established specialties. And the evidence on how thinly that expertise is distributed is consistent:
- Cardio-oncology: only 39 percent of 131 cardiology trainees worldwide had received specific cardio-oncology training (64 percent in the UK, US, and Canada versus 37 percent elsewhere), and roughly a third had no cardio-oncology service at their workplace.
- Obstetric medicine: in one internal medicine needs assessment, 75 percent of staff perceived deficiencies in managing medical problems in pregnancy due to limited exposure. The literature notes that in places without an obstetric physician, informal advice given without ever meeting the patient is routine practice.
- Fertility preservation: roughly 60 percent of primary care physicians in one study lacked adequate knowledge of options, with only 26 to 32 percent aware of relevant guidelines.
- Transgender care: a specialist e-consult service answered in a median of 1.2 days, using 18 minutes of specialist time, avoiding 32 percent of referrals, with 95 percent rating it high value. It was available only to primary care physicians inside one Ontario region. Meanwhile a JAMA Network Open study found 31.7 percent of 489 transgender adults were considering do-it-yourself hormone therapy as access narrows.
These are the fastest-growing areas of medicine, and they are the ones credentials are least able to represent, because a board takes years to establish and the clinical need arrives immediately.
There is one more finding worth highlighting, because it undercuts the assumption that this is a primary care problem. In Ontario's e-consult data, specialist-to-specialist e-consults represented about 4 percent of volume (3,285 consultations) and avoided a face-to-face referral 69 percent of the time.
Specialists are asking other specialists, and it works. The need for cross-boundary expertise routing is not a generalist need. It runs all the way up.
Why nobody indexes the straddler
The clinicians who hold this expertise are real, identifiable, and unfindable. Six reasons:
No board exists for most of it. Cardio-oncology has no US board certification. Neither does obstetric medicine. Without a credential, there is no field for a directory to populate.
The fields are too small to earn boards quickly. Certification processes take years and require a critical mass. Clinical need appears immediately and moves faster.
Self-declared interest is not corroborated. Any clinician can list an interest. Interest and exposure are different things, and nothing distinguishes them.
Neither parent society indexes them. The cardiology society organizes by cardiology. The oncology society by oncology. The cardio-oncologist appears in both and is findable through neither, because each society's taxonomy stops at its own boundary.
The experts are overloaded and therefore invisible by choice. A clinician who becomes known as the regional straddler is inundated. Some deliberately keep a low profile, which is a rational response to unbounded demand with no compensation.
And the nodes are disappearing. With adult congenital heart disease fellowship output failing to offset attrition, and single-specialty adult hematology programs down to two, the graph is losing nodes faster than it adds them.
That last point creates genuine urgency. The remaining experts in several of these fields are aging out. Their exposure, accumulated over decades of managing conditions almost nobody else has seen, is currently recorded nowhere and will simply stop existing.
What the transition literature usually misses
The dominant framing of the pediatric-to-adult transition problem is patient-centered: adolescents need to learn self-management, transition programs should build health literacy, young adults are developmentally prone to disengagement.
All true, and all incomplete.
The measured failures include 42 percent with three-year care gaps and 25 percent lost to follow-up after referral, with patients frequently reporting they did not know follow-up was required. Those are failures of the handoff, not of patient motivation.
A transition is a referral between two clinicians. If the receiving clinician cannot be identified, no amount of patient preparation fixes it.
Reframed as a clinician-discovery problem, the intervention changes entirely. Instead of teaching a nineteen-year-old to advocate for specialist care, you connect the pediatric subspecialist directly to a named adult clinician with relevant exposure and a warm handoff. That is a fundamentally more reliable mechanism, and it is the one nobody has infrastructure for.
What would work
Index by condition-in-population, not by organ and age. "Adults with sickle cell disease." "Adults with Fontan physiology." "Pregnancy in cardiac disease." These are the categories clinicians actually search for and no directory contains.
Attested exposure rather than credentials. Since most of these fields have no board, exposure is the only available signal: how many, how recently, in what setting, corroborated by peers who observed it. This is the only scalable credential for a field that moves faster than certification.
Corroboration from both parent specialties. A cardio-oncologist should be attested by cardiologists and by oncologists, because each sees a different half of the competence.
Explicit transition partnerships. Pairing a pediatric subspecialist with a named adult clinician for a defined caseload, rather than transferring into a directory search. This converts an anonymous handoff into an accountable one.
Capture the retiring experts now. Semi-retired pediatric subspecialists and the few adult experts in contracting fields hold exposure that is about to disappear permanently. Making them routable, with their legacy exposure explicitly marked as historical, is both a valuable service and a race against time.
And handle jurisdiction-sensitive care carefully. For gender-affirming care in particular, any system must allow member-controlled visibility and avoid location-revealing data, given the political and legal environment in some jurisdictions.
What you can do now
If you are an adult clinician
Learn to recognize the category. The question "was this condition diagnosed in childhood?" is a useful trigger. If the answer is yes, standard adult management may not apply, and a surveillance guideline you have never read may exist.
Ask the pediatric side directly. Pediatric subspecialists are frequently delighted to be called about patients they transferred, and they usually remember them. This is the single most available and least used resource in this whole area.
Do not assume the subspecialty covers it. As one patient observed, some hematologists do not deal with sickle cell. Being in the right specialty is not the same as having the relevant exposure, which is exactly the credential-expertise gap in its purest form.
If you are a pediatric subspecialist transferring patients
Transfer to a person, not to a specialty. With 25 percent lost to follow-up and 11 percent failing to transfer at all, a named receiving clinician with confirmed acceptance is the intervention.
Offer ongoing availability explicitly. Telling the receiving clinician they can call you converts a handoff into a relationship, and it is free.
Write down your own exposure before you retire. In contracting fields, you may be one of a small number of people who have managed a given condition at volume. That knowledge currently ends with your career.
If you lead a program or society
Publish an interface roster. Which of your members have genuine exposure to conditions at the boundary with an adjacent specialty. Neither parent society currently does this and both could.
Count your nodes. For any pediatric-onset condition your specialty inherits, how many adult clinicians in your region have real exposure? Most organizations have never asked, and in several fields the answer is alarming.
Frequently asked questions
How many adult congenital heart disease specialists are there? Approximately 509 ABIM-certified adult congenital heart disease cardiologists for roughly 1.5 million US adults with congenital heart disease, or about 500 patients per board-certified physician, with fewer than half practising ACHD full time. Prevalence grows around 5 percent a year and fellowship output does not offset attrition.
What happens to children with chronic conditions when they become adults? Frequently they fall out of specialist care. Research found 42 percent of adult congenital heart disease patients had gaps of more than three years, and in one referred cohort 25 percent were lost to follow-up while 11 percent failed to transfer at all. Common reasons include patients not knowing continued follow-up was required.
Why is adult sickle cell care so hard to access? The infrastructure contracted substantially: single-specialty adult hematology programs fell from 74 in 1995 to two in 2018, and 46 percent of hematologists report a shortage in classical hematology. In one California analysis, 49.1 percent of adult Medicaid patients with sickle cell disease and chronic kidney disease had no hematologist visit over a decade.
What are hyphen specialties? Areas of expertise sitting between established specialties, such as cardio-oncology, obstetric medicine, onco-fertility, and onco-nephrology. They address rapidly growing clinical needs, frequently have no board certification, and are indexed by neither parent specialty's directory, which makes the clinicians who hold the expertise very difficult to find.
Do specialists also need help finding cross-boundary expertise? Yes. In Ontario e-consult data, specialist-to-specialist consultations represented about 4 percent of volume and avoided a face-to-face referral 69 percent of the time, indicating that the need for cross-boundary routing extends well beyond primary care.
How should a pediatric specialist transfer a patient to adult care? To a named individual with confirmed acceptance and relevant exposure, rather than to a specialty or a clinic in general, with an explicit offer of ongoing availability to the receiving clinician. The measured failure rates for anonymous transfers are high, and the mechanism that fixes them is a person on the other end.
The bottom line
Medicine saved these children. That is a genuine triumph and it deserves to be stated as one.
Then it handed them, as adults, to a system organized by organ and by age, which has no category for what they have, no directory field that describes it, and in several cases fewer than a thousand physicians in the country who genuinely know their disease.
Roughly 1.5 million adults with congenital heart disease and 509 boarded specialists, fewer than half full time. Adult hematology programs down from 74 to two. Eighty-five percent of internists missing guideline surveillance for childhood cancer survivors in a vignette study. Nearly half of adult Medicaid sickle cell patients in one state seeing no hematologist at all.
And the remaining experts are retiring, in fields whose training pipelines are shrinking, taking with them the accumulated exposure of careers spent managing conditions almost nobody else has seen.
The transition problem is usually described as teaching young adults to manage their own care. Some of it is. Much more of it is that their new clinicians cannot find anyone who knows the disease, and nothing in medicine is built to help them look.
Part of a series on the missing professional infrastructure of healthcare. Previously: The Nurse You Cannot Find
Evidence note: adult congenital heart disease workforce and outcome figures come from the American Journal of Cardiology (2025), European Heart Journal Supplements (2026), JACC Advances summit proceedings (2026), Gurvitz et al. in JACC (2013), and CJC Pediatric and Congenital Heart Disease (2026). Hematology workforce contraction is reported via STAT (2023). Sickle cell access data comes from JMIR Public Health and Surveillance (2024) and Blood Advances (2022). Hyphen specialty training data comes from Cardio-Oncology (2026), Obstetric Medicine (2026 and 2017), F1000Research (2023), Champlain BASE e-consult evaluation, Journal of Telemedicine and Telecare (2025), and JAMA Network Open (2025). Some prevalence estimates in this area vary by source and definition and should be verified before republication.
