We built hospitals the way we once drew maps. By naming the territories. Cardiology. Nephrology. Gastroenterology. Neurology. Each discipline a sovereign state with its own language, its own protocols, its own definition of a good outcome. The body, in this schema, is a federation of organs. Treat the organ. Discharge the patient.

It is an extraordinary achievement. In isolation, modern medicine is formidable. The cardiologist managing heart failure, the oncologist reading a tumour response, the hepatologist tracking fibrosis. Each at the frontier of what human knowledge can do. The parts are brilliant. The system they compose is something else.

Because biology was never organised this way.

The gut does not observe department boundaries. Microbial metabolites produced in the colon appear in cerebrospinal fluid. Intestinal permeability drives systemic inflammation that surfaces in the joints, the skin, the vasculature, the brain. The microbiome operates through axes that have no equivalent in the hospital org chart: the gut-brain axis, the gut-lung axis, the gut-liver axis. Active, bidirectional, chemically dense. Medicine has spent two decades beginning to characterise them. It has yet to structurally accommodate them.

Nobody owns the axis. That is not a metaphor. It is a description of how hospitals are actually organised.

The gastroenterologist owns the gut. The neurologist owns the brain. The conversation between them, mediated by microbial metabolites, vagal signalling, immune modulation, short-chain fatty acids crossing the blood-brain barrier, belongs to nobody. It falls into the space between departments: between billable codes, between referral pathways, between job descriptions. It disappears into the org chart.

Depression is the leading cause of disability worldwide. Approximately ninety percent of the body’s serotonin is produced in the gut, modulated by microbial activity. A psychiatrist managing a patient with treatment-resistant depression has no structural prompt to ask what is happening in the gut. The system was not designed to surface that question. It was designed to route the patient to the right department, and depression routes to psychiatry.

Rheumatoid arthritis, inflammatory bowel disease, Parkinson’s disease, cardiovascular disease, metabolic syndrome: the literature connecting each of these to microbial community structure and metabolic output is no longer speculative. It is peer-reviewed, replicated, and growing faster than clinical practice can absorb it. The science is maturing. The system receiving it was built for a different model of the body.

The cardiologist is not ignoring the gut-heart axis out of negligence. This is not a failure of individual clinicians. They are operating within an architecture that does not ask them to look there, does not train them to look there, does not reimburse them for looking there. The failure is structural. The incentives are misaligned at the level of institutional design, not individual competence.

Institutional design is slow. What medicine has built over a century and a half does not reorganise around a new paradigm because the paradigm is correct. It reorganises when the cost of not reorganising becomes impossible to absorb, financially, clinically, politically. We are not there yet. The gaps are real. They show up as readmissions, treatment resistance, patients with complex multimorbidity who fall between specialties and never quite get better, and costs that accumulate in the spaces the org chart cannot see.

The microbiome did not create this problem. It revealed it.

What the microbiome makes visible is that the body is not a federation of organs. It is an ecosystem. Ecosystems do not respond to interventions aimed at components in isolation. They respond to changes in conditions, in relationships, in the flows between parts. Treat the gut in isolation and you miss what the gut is doing to the brain. Treat the brain in isolation and you miss what is driving the inflammation. The unit of analysis is wrong. When the unit of analysis is wrong, even excellent science, applied with precision and good intent, lands in the wrong place.

The question for the next decade is not whether the microbiome is clinically relevant. That question is settled. The question is whether institutions built to deliver healthcare can reorganise around a biology they did not anticipate, one that is relational, systemic, and indifferent to the departmental boundaries medicine drew for its own convenience.

Biology never got the memo.

It has been connecting, signalling, adapting, failing across every boundary we have drawn. It was doing this long before we had the tools to see it. We are beginning to see it now. The harder work is building systems capable of acting on what we see.

That work is institutional. It is political. It is economic. The question is whether medicine is willing to be wrong about its most foundational assumption: that the body is best understood in parts.

It is not. It never was.