The Last Mile of Health Care: Why Discounts Only Matter When They Become Relationships

What if the biggest obstacle to better health care is not a lack of medical knowledge, but a failure to connect money, institutions, and daily life?

America spends enormous sums trying to make care more accessible and more effective. Yet a discount on medicine may remain trapped inside an institution, while a sophisticated care model may struggle to reach the patient’s home. The missing ingredient is not always another program. It is often a conversion system, a way to turn financial resources into local capacity, local capacity into continuity, and continuity into better health.

This is the deeper connection between two seemingly separate developments in American health care: the rapid scale of the 340B drug purchasing program and the rise of community based, technology enabled care for people with chronic kidney disease. One concerns the price institutions pay for medicines. The other concerns how physicians and communities organize care. Together, they raise a more consequential question:

When health care receives resources, what determines whether those resources become care that patients can actually feel?

The difference between funding care and building care

The 340B figures are striking. Covered entities purchased approximately $53.7 billion in medicines through the program in 2022. Disproportionate share hospitals accounted for about $41.8 billion, nearly four fifths of the total. Health center programs accounted for approximately $2.8 billion, while children’s hospitals, rural referral centers, critical access hospitals, clinics serving people with HIV, and other safety net organizations made up the remainder.

These numbers measure purchases, not necessarily the amount of assistance patients received, the discounts retained, or the health outcomes generated. That distinction matters. A financial resource is not the same thing as a care capability. A discount can improve an organization’s finances without automatically improving the experience of a person who cannot get an appointment, lacks transportation, misunderstands a prescription, or is discharged without reliable follow up.

Consider two hospitals receiving the same effective financial advantage. The first uses it mainly to stabilize a complex institution with expensive infrastructure. That may be socially valuable. The second invests in medication access, community health workers, home visits, patient education, and coordination with primary care. Both may be serving vulnerable populations, but they are converting resources through different mechanisms.

The key issue is therefore not simply whether money enters the health system. It is whether the system has a reliable last mile for turning money into patient benefit. In health care, the last mile is not a delivery truck. It is the moment when a patient understands what to do, can obtain the treatment, has someone to call, and receives help before a manageable problem becomes an emergency.

This is especially important in chronic kidney disease. Kidney disease rarely behaves like a single episode that begins and ends in a hospital. It unfolds over time through blood pressure, diabetes, medication choices, nutrition, transportation, housing, family support, and repeated clinical decisions. The disease is biological, but its trajectory is also logistical.

A system that pays for the right drug but does not help a patient take it consistently has solved only part of the problem. A system that employs excellent specialists but cannot reach a patient between appointments has also left value on the table.

Chronic illness exposes the boundaries of the clinic

Traditional clinical practice is organized around encounters. A patient enters a clinic, receives an evaluation, gets a prescription or recommendation, and leaves. This model works reasonably well for some acute problems. It is poorly matched to illnesses in which the most important determinants of outcome occur outside the exam room.

A person with advanced kidney disease may need help understanding dietary restrictions, arranging transportation, coordinating multiple specialists, identifying worsening symptoms, managing medications, and deciding whether to call a physician or go to the emergency department. Family members may be central to the plan. Community organizations may know more about a patient’s practical barriers than any clinical record does.

That is why a community based care model is more than a change in location. It is a change in the unit of care. The unit is no longer the individual appointment. It becomes the network around the patient: clinicians, family, technology, home based services, community resources, and timely communication.

A technology platform can help identify risk, but risk identification is not intervention. An algorithm may flag a patient whose kidney function is deteriorating. The patient still needs a person who can reach them, understand why they are struggling, and coordinate the next step. Technology is valuable when it shortens the distance between noticing a problem and doing something about it.

The same principle applies to physician participation. A value based model can preserve physician independence while providing shared infrastructure for engagement, data, and home based care. This matters because specialization and coordination are not opposites. A nephrologist should not have to become a transportation coordinator, social worker, data analyst, and after hours triage service simply to practice excellent medicine. But without those functions, excellent medicine may arrive too late.

The most effective model is therefore neither purely institutional nor purely individual. It is distributed but coordinated. Responsibility is placed close to the patient, while information and support move across organizational boundaries.

The future of chronic care will belong to systems that make the patient’s environment part of the clinical team.

The hidden economics of the last mile

The connection between drug purchasing and community care becomes clearer when we think in terms of a simple chain:

Resource to capacity to contact to continuity to outcome.

A discount, grant, or reimbursement arrangement is a resource. It becomes capacity when an organization uses it to hire people, build data systems, extend hours, or create partnerships. Capacity becomes contact when someone can reach the patient in a meaningful way. Contact becomes continuity when the relationship persists across settings and over time. Only then can better outcomes reliably emerge.

At every step, value can be lost.

A hospital may have financial resources but lack a way to reach patients at home. A community clinic may have trusted relationships but insufficient specialty support. A physician group may have clinical expertise but no data infrastructure. A patient may receive a care plan that is medically sound but impossible to follow given their transportation, work schedule, or family responsibilities.

This suggests a useful management concept: the care conversion ratio. It asks how much of every dollar, data point, or clinical interaction is converted into a patient action that improves health.

The ratio rises when organizations:

• Put resources close to the barriers patients actually face.
• Share information across clinicians and settings.
• Give patients one accountable pathway for help.
• Measure avoided crises, not just completed encounters.
• Combine professional expertise with family and community knowledge.

The ratio falls when organizations:

• Count purchases as outcomes.
• Count appointments as engagement.
• Count referrals as coordination.
• Count software installation as digital transformation.
• Treat social barriers as someone else’s responsibility.

This framework does not imply that hospitals are wasteful or that discounts are ineffective. It does imply that institutional scale and patient proximity are different kinds of value. A large institution may be excellent at purchasing, compliance, acute treatment, and complex procedures. A local care network may be better at trust, persistence, and adapting plans to daily life. A resilient system needs both, but it must connect them deliberately.

Why scale can solve one problem and create another

The 340B purchasing data also reveal a structural tension. The program reaches many types of covered entities, including rural hospitals, health centers, children’s hospitals, cancer centers, tribal health programs, and clinics serving people with HIV. Yet the largest share of purchases belongs to disproportionate share hospitals.

Scale brings advantages. Large organizations can negotiate, operate complex pharmacies, maintain compliance departments, and spread fixed costs across many patients. But scale can also make it harder to see where value lands. A dollar moving through a large institution may support important functions while becoming difficult to trace to a specific patient experience.

This is not an argument for abandoning scale. It is an argument for pairing scale with granular accountability. The question should not merely be, “How much value did the institution capture?” It should also be, “What new patient facing capability did that value create?”

For example, an organization might report that its resources supported a care team for high risk kidney patients. A stronger account would show what the team changed: how quickly patients were contacted after a hospital discharge, how many medication problems were resolved, how often urgent visits were prevented, whether patients received care at home, and whether nephrologists had more time for decisions requiring specialist judgment.

Metrics should follow the actual causal chain. If the goal is better kidney care, relevant measures might include:

• Time from a concerning lab result to patient contact.
• Percentage of patients with a documented, understandable care plan.
• Medication access and adherence barriers resolved.
• Avoidable emergency visits and hospitalizations.
• Patient reported confidence in managing symptoms.
• Whether family members and community resources were incorporated when appropriate.

These measures are not merely administrative details. They reveal whether the system is functioning as a bridge or as a collection of disconnected platforms.

Designing health care as a network, not a warehouse

The conventional health system often resembles a warehouse. Resources are stored inside institutions: medicines in pharmacies, expertise in specialist offices, data in separate software systems, and social support in agencies with limited clinical visibility. Patients must navigate between these stores, carrying their own information and coordinating their own movement.

A better design resembles a network. Resources remain distributed, but the connections are intentional. A nephrologist can see relevant information. A care team can make a home visit. A community provider can flag a practical barrier. A family member can participate with the patient’s permission. The system does not require every organization to do everything. It requires each organization to make its contribution legible and reachable.

This is where physician led partnerships become important. Local specialists bring trust, judgment, and knowledge of the communities they serve. A supporting organization can provide technology, patient engagement methods, and operational capacity. The partnership works when it amplifies local expertise rather than replacing it.

The same design principle can guide the use of 340B related resources. Instead of treating purchasing power as an endpoint, organizations can ask which shared capabilities would improve the entire local network. Examples include a regional medication navigation service, a common referral and follow up process, mobile clinical teams, shared care coordinators, or a data system that alerts clinicians when high risk patients move between hospital and home.

The important idea is shared infrastructure for continuity. No single physician, hospital, or clinic can solve every barrier. But a network can make it much less likely that a patient falls through the gaps between them.

Key Takeaways

1. Separate financial inputs from patient outcomes. A large purchasing figure shows resource flow, not necessarily health impact. Track how resources become staffing, access, continuity, and measurable patient benefit.

2. Design around the patient’s real environment. For chronic disease, include transportation, family support, medication access, housing, work, and communication in the care plan.

3. Build a care conversion chain. For every new dollar, tool, or partnership, ask: What capacity does this create? Who will contact the patient? What happens next? How will we know the intervention worked?

4. Pair institutional scale with local trust. Large organizations can supply purchasing power and infrastructure. Community clinicians and families often supply context, persistence, and credibility. Neither is sufficient alone.

5. Measure the last mile. Track the time from risk detection to action, the resolution of practical barriers, and the prevention of avoidable crises, not only visits, purchases, or referrals.

The most important health care resource may not be money, medicine, or data in isolation. It is the ability to connect all three to a person before the next crisis occurs.

The real test of a value based system is therefore not whether it has discovered a more sophisticated payment model. It is whether a patient with a complicated illness experiences less friction, more understanding, and earlier help. That requires a different mental model of reform. Health care is not improved simply by adding resources to institutions. It improves when institutions become capable of transmitting those resources through trusted relationships and into daily life.

A discount can lower the price of treatment. A care network can make treatment possible. The best systems will understand that these are not separate achievements. They are two parts of the same public obligation: converting society’s investment in health into help that reaches a human being in time.