During my recent stay in Lima, I spent more time than usual accompanying family members through clinics, consultations, tests, and hospital visits.

The experience was not remarkable because medicine failed as a science. In most rooms, the opposite was true. A specialist asked a specific question, looked at the relevant evidence, and made a decision within a field they understood far better than I did.

What became visible was the space between those decisions.

This felt familiar for reasons that had very little to do with medicine. Many years, during my time at R/GA, the phrase we used was “functional integration.” It was the agency’s language for how capabilities that traditionally worked separately were supposed to function as one connected system around the consumer. My own role increasingly sat between client teams, agency disciplines, and outside partners. I was not thinking about healthcare. What stayed useful was the operating problem: expertise can be distributed while the experience still has to be continuous.

A cardiologist looks at one set of signals. Another specialist looks at another. A test answers a specific question. A medication belongs to a particular problem. Each appointment has a reason, a vocabulary, and a defined responsibility.

The patient leaves carrying all of them.

Specialization Solves One Problem and Creates Another

It is easy to criticize fragmented healthcare and harder to propose a serious alternative that does not involve specialization. We want clinicians to know their domains deeply. Modern medicine is too complex for every practitioner to hold every relevant body of knowledge at equal resolution.

The difficulty begins when the boundaries needed to produce expertise become boundaries the patient has to cross repeatedly. One clinician changes a medication. Another needs to know. A new symptom appears. Is it part of the original condition, a side effect, a separate problem, or simply a body under strain? Someone has to remember what changed, when it changed, what was tried, and which detail might matter at the next appointment.

Specialization does not become fragmentation because different clinicians know different things. It becomes fragmentation when the work required to connect those things has no reliable owner.

Frequently, the patient or family becomes that owner.

A systematic review of chronic illness care found associations between greater fragmentation and outcomes including more emergency visits, diagnostic testing, and healthcare costs, while also finding substantial variation across studies. The problem is not that medicine divides complexity. It is what happens when the divisions become weakly connected.

The Handoff Is Where Uncertainty Accumulates

Every specialist sees the patient through a particular question. That is necessary and incomplete by design. The consultation asks what matters for this organ, condition, medication, symptom, or procedure. The patient carries a different question: what is happening to me?

A recommendation can be entirely reasonable inside one domain and still create difficulty somewhere else. A medication may improve one variable while affecting appetite, energy, balance, sleep, or another treatment. A procedure can solve one problem while temporarily changing movement and independence. Advice to rest can collide with the practical need to remain mobile.

No one needs to have made a mistake for the total experience to become difficult to integrate. The uncertainty does not disappear when a patient moves from one specialist to another. Some of it simply changes owner.

That change is easy to miss because the handoff often looks administrative. Someone has to carry the updated medication list, notice that dizziness began after a dose changed, check whether one specialist has seen another’s result, reconstruct which symptom came first, and prepare the question that connects two appointments never designed as one conversation.

The Family Carries Continuity

Caregiving contains a great deal of work that does not look medical until it goes wrong. Appointments have to be scheduled, results retrieved, medication lists updated, instructions remembered, questions prepared, transport arranged, and symptoms noticed between visits. Someone has to know which change happened before which other change.

Medication management alone can be substantial. A 2025 systematic review of informal caregiving documented family members performing practical tasks such as obtaining and administering medicines alongside cognitive work such as decision-making and interpretation. That description matched what I kept seeing around me.

The clinical record contained facts. The household contained continuity.

Family members see what happens after the medication is taken: whether sleep changes, appetite disappears, the person becomes weaker or more confused, or a routine quietly stops working. They see the kitchen, the stairs, the missed meal, the reluctance to go outside, and the conversation that makes no sense until someone remembers what happened three days earlier.

None of this turns a relative into a clinician. It gives them access to a different kind of evidence: what the patient is like between episodes of care.

The formal system is organized around encounters. The household sees the transitions between them.

The Body Keeps Integrating What Institutions Separate

The body does not honor the categories used to organize care. Pain can reduce movement. Reduced movement can change confidence and independence. That can alter mood, social contact, appetite, sleep, and willingness to follow a treatment plan. The sequence may begin in one department and become visible in another.

The same problem appears in quieter forms. A week of appointments changes routine and meals. Fear of falling changes movement before measurable decline in strength appears. A poor night changes how someone answers questions the next morning. The categories organize expertise, but the effects keep crossing them after the appointment ends.

A reassuring laboratory result does not prove that the person is functioning well. Feeling reasonably well does not prove that an important clinical problem is absent. A family member saying, “She has not been herself since Tuesday,” is not a diagnosis, but it can still matter because it carries a time marker and a deviation from baseline.

The question is whether there is somewhere for that observation to meet the formal evidence.

Coordination Is Not Administrative Overhead

We often picture care as the direct clinical act: examination, diagnosis, prescription, procedure. Everything around it can look secondary: scheduling, handoffs, medication reconciliation, discharge instructions, communication across specialties, follow-up.

But the effect of a clinical decision depends on that chain. Does the next clinician know what changed? Does the patient understand the treatment? Has the relevant result reached the person making the next decision?

If a medication list is wrong or a change is never carried forward, the failure can look administrative until it alters care.

Much of coordination looks administrative because its objects are ordinary: an appointment, a result, a medication list, a phone call. But if the effectiveness of one clinical decision depends on information surviving the journey to the next one, continuity is part of the care itself.

The more complex a patient’s life, the more consequential this becomes. Complexity is not only the number of diagnoses. It is also the interactions among diagnoses, treatments, physical limits, daily routines, and the people helping manage them. A patient with three well-managed conditions may have a simpler course than someone with one condition whose treatment changes movement, sleep, appetite, independence, and the household around them.

Medical complexity and lived complexity are related, and not interchangeable.

Whole-Person Care Does Not Require an Omniscient Clinician

“Whole-person care” can become vague quickly. It cannot mean that every specialist should investigate every dimension of a patient’s life at every appointment. That would weaken the specialization that may be required.

The more practical requirement is continuity. A local decision enters a person who already has other treatments, physical constraints, routines, resources, relationships, and people helping them manage what happens next.

The specialist does not need to own all of that. But enough of it has to travel with the patient for the decision to remain intelligible inside the larger course of care.

Sometimes primary care provides that integration. Sometimes a care team does. Sometimes a family member notices that two pieces of information belong in the same conversation.

The mechanism matters more than the label.

The Patient Lives in the Interval

Most of a health trajectory happens outside a consultation room. It happens after the patient gets home: when the medication is taken, the stairs have to be climbed, appetite changes, an instruction proves difficult to follow, or someone notices that the person who came home on Tuesday is behaving differently by Thursday.

These details can sound softer than laboratory values because they are harder to quantify. They are also where many clinical decisions become real. The specialist sees the bounded problem at high resolution. The patient and household live what happens when that decision encounters everything already in motion.

A few weeks ago I wrote that no one lives in pieces. Healthcare makes the point very concrete. Departments are necessary. So are specialists, tests, and bounded questions. The patient still has to live their combined consequences as one life.

When continuity works, specialization becomes more powerful because information can travel with the person. When it does not, the patient and family become the place where fragmentation accumulates.

That is less an indictment of individual doctors than a design problem. Institutions divide complexity so they can act on it. Human beings still have to live the whole.

If this helped you name something you have been feeling but had not quite put into words, please share it with someone who might need the same language.

The Human OS Manual is now . If this work resonates, you can reserve your copy and follow the ideas here as they continue to develop.

Feel free to leave a comment. I read and respond to every thoughtful note.

Notes and Sources

Jee Young Joo, “Fragmented Care and Chronic Illness Patient Outcomes: A Systematic Review,” Nursing Open (2023).
Maha Alkhaldi, Laura Lindsey, and Charlotte Richardson, “Role of Informal Carers in Medication Management for People with Long-Term Conditions: A Systematic Review,” BMJ Open (2025).
Campaign, “R/GA’s Prescription for Future Agencies: Build Business Ecosystems” (2014).