The Hidden Economics of Better Care: Why Friction, Not Information, Determines What Health Systems Deliver

What if the biggest waste in health care is not a lack of information, but the effort required to move it to the right place at the right moment?

A patient may have coverage for a lower cost medication but be unable to compare it easily. An employee may need therapy but encounter a benefits maze that makes the first appointment harder than an emergency department visit. A clinician may know that a patient has a plan, a network, and a treatment history, yet still lack a practical way to assemble those facts into a useful decision.

These are often treated as separate problems: one belongs to technology, the other to population health. They are connected by a deeper principle. Health care systems do not merely provide choices. They shape the routes people take through care. When those routes are difficult to navigate, people select the most visible or immediately available option, not necessarily the most effective one.

The central question, then, is not simply whether data are available or whether a service is covered. It is this: Can the system reduce the friction between a person’s need and the intervention most likely to meet it?

Information Has Value Only When It Changes a Route

Health care has spent decades treating information as a scarce resource. The solution has often been to collect more of it: more records, more claims, more directories, more benefit details. But information sitting inside a portal, database, or disconnected workflow is not yet useful. It becomes valuable only when it changes what someone can do.

Consider a person choosing a prescription. A formulary API can expose which drugs a payer covers and under what conditions. A plan network API can show which clinicians and facilities are available. A consumer directed data API can allow an individual to bring payer information into an application of their choice. Each capability reduces a different kind of uncertainty: price, access, or personal control.

But an API does not automatically produce better care. It is more like a road system than a destination. A new road matters when it makes a beneficial destination easier to reach, not merely because asphalt has been added. If the route is poorly marked, disconnected from other roads, or available only to specialized vehicles, the nominal improvement may have little practical effect.

This is why testing matters. Open, reusable tests for payer APIs do more than check technical compliance. They establish whether the roads connect in predictable ways. A developer should not have to guess how a network directory represents a clinician, or whether a formulary response can be interpreted consistently across plans. Shared tests turn interoperability from a slogan into an observable property.

Yet technical connection is only the first layer. The harder question is what behavior the connection enables. If a patient can see a lower cost medication but cannot understand whether it is clinically appropriate, the system has transferred information without transferring decision making. If an application can retrieve a directory that is technically valid but practically outdated, the user is given a map whose roads may no longer exist.

The real unit of interoperability is not the data exchange. It is the successful decision that the exchange makes possible.

This distinction becomes especially important when we examine how people use mental health services.

The Most Expensive Path Is Often the Path of Least Resistance

A workforce mental health program can appear, at first, to increase spending. Participants may attend more therapy sessions than people using the conventional medical benefit. That observation alone could lead an organization to conclude that the program is generating additional utilization.

A broader view produces a different interpretation. Participants used the dedicated program for roughly twelve sessions per person per year, compared with about seven mental health office visits under the medical plan among nonparticipants. At the same time, spending on the medical plan and prescription drugs was lower for program users, and mental health related emergency department visits also declined.

The important result is not simply that therapy can save money. The deeper result is that changing the route into care changed the pattern of care itself. A service that was easier to access, more specialized, or better matched to the need redirected people away from more expensive channels.

This is a general systems pattern. When a low friction option is absent, people do not stop needing care. They improvise. They wait until symptoms worsen. They use a primary care visit to address a problem requiring behavioral expertise. They seek emergency help because it is available at any hour. They pay out of pocket, abandon the search, or rely on informal advice.

The eventual cost is then attributed to the illness, but some of it was created by the pathway. A person who enters care late may require more intensive treatment. A worker who cannot find an in network therapist may miss work or turn to an emergency department. A patient who cannot determine whether a medication is affordable may stop taking it, leading to complications that appear later in the claims data.

This suggests a useful mental model: care utilization is partly a map of system friction. High use in an expensive setting does not necessarily mean people prefer that setting. It may mean the system has failed to make a more appropriate setting visible, available, trusted, or navigable.

The mental health example also introduces a subtle measurement problem. If an intervention works, the wrong metric may make it look wasteful. Counting therapy visits alone misses the reduction in medical spending, medication spending, and emergency utilization. Likewise, counting API transactions alone tells us little about whether people found an appropriate clinician, selected a viable medication, or gained control over their health information.

The relevant question is not, “Did utilization increase?” It is, “Did the distribution of utilization become more appropriate?”

From Data Portability to Care Portability

The connection between payer interoperability and mental health economics becomes clearer if we distinguish two kinds of portability.

Data portability means that information can move. A person can retrieve coverage details, a developer can query a formulary, and an application can access payer data with permission. This is necessary because fragmented information creates avoidable search costs.

Care portability means that a person can use that information to move smoothly toward the right service. It includes the ability to identify an available clinician, understand the likely financial consequences, compare options, schedule an appointment, and carry relevant information across organizations.

Data portability is infrastructure. Care portability is an outcome.

Imagine a traveler with a digital map that displays every road but does not indicate closures, tolls, traffic, or whether the destination is open. The map is technically comprehensive but operationally weak. Health care directories and benefit systems often have the same problem. They expose fields while leaving the person to perform the integration.

A genuinely useful ecosystem would connect the steps. A person looking for mental health support might authorize an application to retrieve plan coverage, identify clinicians in the relevant network, display appointment availability, clarify costs, and share necessary information with the chosen provider. The system would not merely reveal the existence of care. It would reduce the number of decisions and searches required to reach it.

This is where standards, testing, and program design meet. Standards make information structurally comparable. Testing makes the structure dependable. Program design determines whether the information is placed inside a workflow that people can actually use. The three layers are mutually necessary.

Without standards, every connection becomes a custom negotiation. Without testing, standards remain aspirational. Without a useful workflow, reliable data becomes another administrative burden.

The same framework applies to formulary and network information. Suppose a patient’s application shows that a medication is covered, but does not explain prior authorization requirements, estimated cost, or nearby clinicians who can prescribe it. The information may satisfy an exchange requirement while failing the patient’s real objective. The system has optimized the transmission of facts instead of the completion of a task.

This leads to a more demanding definition of interoperability: interoperability is the capacity of separate systems to cooperate around a human goal. The goal might be starting therapy, selecting an affordable medication, finding an in network specialist, or moving records to an application that helps a person manage care.

The Friction Budget: A Practical Way to Design Better Systems

Every care journey has a friction budget. People will tolerate only so many searches, confusing forms, uncertain prices, delayed responses, and repeated explanations before they abandon the process or choose a less suitable alternative.

Friction is not always visible as a line item. It appears as time spent calling benefits offices, failed appointments caused by inaccurate directories, duplicated assessments, untreated symptoms, and emergency visits that might have been avoided. Organizations often measure the financial cost of the final service while ignoring the invisible cost of navigating toward it.

A useful design process begins by mapping the route rather than the department. Take the journey of an employee seeking therapy:

1. The employee recognizes a problem.
2. The employee decides whether the problem is serious enough to justify help.
3. The employee learns what benefits exist.
4. The employee finds a clinician who is available and covered.
5. The employee estimates cost and privacy implications.
6. The employee schedules an appointment.
7. The clinician receives enough information to begin effectively.
8. The employee continues care when appropriate.

A breakdown at any point can redirect the person into a more expensive path. A benefit may exist but be hard to discover. A directory may list clinicians who are not accepting patients. A program may offer therapy but fail to communicate how it relates to the medical plan. A person may receive care but have to repeat the entire story when transitioning between services.

The same route based analysis can be applied to payer APIs. For each data element, ask four questions:

• Who needs this information?
• What decision should it support?
• What action becomes possible after the decision?
• How will we know that the action improved the journey?

These questions prevent a common failure mode: mistaking technical completeness for practical usefulness. A system may return every required field and still impose too much friction on the person who needs to act.

Organizations should also measure route substitution, not merely volume. Did more people use an evidence based therapy program while fewer relied on emergency care? Did clearer formulary information reduce abandoned prescriptions? Did accurate network data shorten the time from search to appointment? Did consumer access to payer data lead people to use tools that improved adherence or continuity?

These are harder questions than counting visits or API calls, but they are closer to value.

Building Systems That Redirect Behavior on Purpose

The emerging lesson is not that every digital tool will reduce cost, or that every mental health program will produce savings. It is that well designed access can alter the economics of care by changing the default route.

That requires several disciplines to work together.

First, treat interoperability as a product experience, not just a compliance function. A technically valid API should be evaluated by the quality of the decisions and actions it supports. Testing should eventually include realistic use cases, error handling, data freshness, and the ability to move from information to appointment or treatment.

Second, invest in the middle layer between raw data and clinical action. People need interpretation, context, and next steps. A formulary response should help users understand options. A network directory should help them find someone they can actually see. A consumer data connection should support continuity rather than simply create another data silo.

Third, evaluate programs across the whole system. If therapy use rises while emergency visits and medical spending fall, the additional therapy is not necessarily an inefficiency. It may be a substitution toward a more appropriate form of care. Any serious evaluation should examine total spending, avoidable escalation, access time, continuity, and user experience together.

Finally, design for the person who is least able to absorb friction. People in distress have less patience for confusing interfaces. People with limited time cannot call five offices to verify a directory. People managing multiple conditions cannot repeatedly reconstruct their history. Reducing friction is not merely a convenience improvement. It is a form of equity because the burden of navigation falls most heavily on those with the fewest resources.

Key Takeaways

• Measure completed journeys, not just exchanged data. Ask whether information helped a person choose, schedule, start, or continue appropriate care.
• Look for route substitution. More use of an accessible, evidence based service may be positive if it reduces emergency care, avoidable medical utilization, or treatment abandonment.
• Connect coverage information to action. Formulary and network data should lead toward understandable choices, realistic availability, and clear next steps.
• Test the full experience. Technical conformance is necessary, but data freshness, interpretability, workflow integration, and error recovery determine practical value.
• Treat friction as a measurable cost. Track time to appointment, repeated contacts, abandoned searches, duplicated assessments, and transitions between benefits.

The future of health care interoperability will not be decided by how many systems can exchange messages. It will be decided by whether those exchanges make better routes more obvious and easier to follow.

A person rarely experiences an API, an implementation guide, or a claims category. They experience a moment of uncertainty: Which medication can I afford? Who can see me this week? Where do I go before this becomes a crisis? The systems that answer those questions with the least friction will shape behavior long before anyone notices the architecture underneath.

The best health care infrastructure does not merely make information move. It makes appropriate care easier to choose than inappropriate care.

That is the hidden economics of access. Better information matters, but only when it changes the path a human being takes.