The Sequencing Is Easy. The Hard Part Is Everything After.

In the two decades since the Human Genome Project was completed, the cost of reading a person's DNA has fallen faster than almost any technology in history. What once required a global consortium and roughly three billion dollars can now be done in a clinical lab for the price of an MRI. And yet, for the majority of Americans, genomic information remains effectively invisible at the point of care. A patient can carry a variant that meaningfully changes their cancer screening schedule or medication dosing — and it will never appear in front of the physician making those decisions.

The sequencing is, by now, the easy part. The hard part is everything that happens afterward.

Records That Cannot Speak the Language

The core problem is that the American electronic health record was built for a different era of medicine. EHR systems are engineered around billing codes, lab values, and medication lists — discrete, standardized data points. Genomic results do not fit neatly into that architecture. A single genome can contain millions of variants relative to a reference sequence, and each of those variants carries meaning only in context: which lab found it, which reference build was used, what classification was assigned, and when.

Efforts to standardize this information exist. HL7's FHIR framework includes a genomics module, and organizations such as the Clinical Genome Resource work to maintain consistent variant interpretation. But adoption is uneven, and most hospital systems in the United States still store genetic results as scanned PDFs or free-text notes buried in the chart. A PDF cannot trigger a clinical decision alert. A scanned document cannot update itself when a variant of uncertain significance is reclassified as pathogenic — a process known as variant reclassification, which happens routinely as evidence accumulates.

The consequence is a kind of clinical amnesia. A result ordered five years ago may be dangerously out of date, and nothing in the record will say so.

The Reimbursement Gap

Even when data flows, the people who can interpret it are often not available — or not paid for. Genetic counselors are the professionals specifically trained to translate genomic findings into medical decisions and family planning conversations. Yet reimbursement for their services remains inconsistent across states and insurers. Medicare, which sets the tone for much of American coverage policy, restricts genetic counselor billing in ways that push counseling into physician visits or leave it out entirely.

The math is unforgiving. There are only around 5,000 certified genetic counselors practicing in the United States against a population of more than 330 million. Without sustainable payment models, scaling genomic medicine through this workforce alone is impossible. Some systems are experimenting with tiered models — chatbot triage, group counseling sessions, physician extenders — but the financing structure has not caught up to the demand that cheap sequencing has created.

A Training Deficit Decades in the Making

The third barrier is human. Most Americans receive their primary care from physicians who trained before routine genomic testing existed. Surveys of primary care providers have repeatedly found low confidence in interpreting genetic results, uncertainty about when to order testing, and discomfort discussing incidental findings. This is not a failure of individual physicians; it is a failure of medical education. Genomics remains a thin slice of most medical school curricula, and continuing education has not closed the gap.

The result is a system in which the people closest to patients are the least equipped to act on one of the most consequential categories of medical data.

What a Genome-Literate System Would Require

Fixing this does not require a scientific breakthrough. It requires infrastructure, and that is in some ways a harder political problem. A genuinely genome-literate health system would include:

• Structured, computable genomic data in every EHR, with automatic reclassification alerts pushed to treating physicians.
• Reimbursed genetic counseling at scale, including telegenetics services that reach rural and underserved communities.
• Genomics competencies embedded in medical training and board certification, not treated as a subspecialty luxury.
• Interoperability standards enforced with the same seriousness as billing standards, so a result from an academic center travels with a patient to a community hospital.

None of these steps is glamorous. All of them are achievable. The scientific community has, in effect, already done its part: the reading is fast, cheap, and increasingly accurate. What remains is the unglamorous work of pipes, payments, and professional education — the difference between having a genome and being able to use one.