One MRI. Three Operating Models.
A physician orders an MRI. Payer, Provider and Patient each see it through a different Operating Model:
Three active participants enter the decision with different information, economics, responsibilities and priorities, and each may make a perfectly rational decision within its own Operating Model.
But who is operating the whole?
BlueHour applies Operating Model Modernization and Operating Architecture to healthcare. The demonstration uses synthetic data, is not medical advice and does not replace licensed clinical judgment.
In healthcare, the governed outcome is Positive Patient Outcomes and Experiences.
Positive Patient Outcomes and Experiences
Every healthcare interaction produces an outcome, and an outcome can be positive or negative. The objective is not simply an outcome. It is to increase the probability of a positive one, and to reduce the probability and consequences of a negative one.
A patient can have a clinically positive outcome while experiencing unnecessary delay, administrative confusion, excessive cost, poor communication or fragmented care. Clinical outcome and patient experience are related but not synonymous. Operating Architecture considers both.
Healthcare does not suffer from a lack of intelligence
Intelligence is everywhere in healthcare. The problem is getting all of it to operate together toward a governed positive outcome.
Payer, Provider and Patient each enter the decision with different information, economics, responsibilities and priorities. The Patient is not simply the recipient of decisions made by the other two. Their authority and responsibilities are not equivalent, but all three contribute to the operating outcome. Each can act rationally inside its own Operating Model while the overall system still produces a result that is slower, more expensive or worse for the patient than it needed to be.
BlueHour addresses the behavior of the whole. AI is one form of intelligence operating inside a much larger healthcare system, and the agent is not the system.
A seemingly simple order becomes a multi-party operating system
Fifteen steps cross three Operating Models, with six handoffs between Patient, Provider and Payer. Each participant is generally doing its job within the system it has been given. The problem is architectural: nobody is assigned the behavior of the whole flow, so nobody is accountable for whether it produces the best possible overall result.
Swipe sideways to follow the whole flow.
The same MRI, seen from three Operating Models
Provider economics are real, but so are clinical responsibility, capacity, physician efficiency, quality and patient care. Payers carry legitimate economic, clinical, risk and administrative responsibilities. Every question below is legitimate. None of them, on its own, describes the behavior of the whole.
- Do I need this MRI?
- Where should I go?
- How quickly can I get it?
- Is the facility appropriate for my clinical needs?
- Is it in network?
- What will it cost me?
- Has it been authorized?
- How far do I have to travel?
- What happens after the MRI?
- When will my physician receive the results?
- When will I know what those results mean?
- Clinical appropriateness
- Diagnostic requirements
- Urgency
- Imaging capability
- Available capacity
- Scheduling
- Quality
- Network considerations
- Prior authorization
- Existing diagnostic information
- Physician workflow
- Results delivery
- Follow-up care
- Administrative cost
- Asset utilization
- Continuity of care
- Medical necessity
- Appropriate utilization
- Prior authorization
- Network status
- Contracted rates
- Alternative facilities
- Duplicate procedures
- Quality
- Patient benefit design
- Patient out-of-pocket exposure
- Total medical cost
- Downstream cost
- Administrative expense
The technology is already there
Healthcare already operates on an enormous technology ecosystem: MRI systems, EHRs, imaging platforms, scheduling applications, payer systems, networks, clinical software and AI. Each performs an important function and generates valuable intelligence. The operating problem is larger: how does all of that technology operate together across Payer, Provider and Patient? BlueHour does not require these systems to disappear. Operating Architecture gives their capabilities and intelligence a way to participate in the larger Operating Model.
BlueHour does not require rip-and-replace. Existing technologies can participate in the Operating Model while their continued role is governed by clinical, operating and economic requirements.
Conceptual. The demonstration shows where this intelligence could come from; it does not claim that any of these integrations exist today.
Healthcare can have a better MRI, a better EHR, a better AI model, a better scheduling application, a better payer platform and a better clinical application, and still produce a poor operating outcome when they interact. The agent is not the system. Neither is the MRI, the EHR or any individual application. The behavior of the whole matters.
Imaging manufacturers such as Siemens Healthineers, GE HealthCare and Philips, and EHR platforms such as Epic, already provide sophisticated imaging, software, workflow and analytics capabilities, and their work extends well beyond the machine. Technology providers optimize important components and capabilities. Operating Architecture works at a different level: how those capabilities participate in the behavior of the larger system across Payer, Provider and Patient. They are part of the system, and BlueHour does not need to replace them. Names are examples of the ecosystem only; no partnership or integration is implied.
Jane Smith needs a lumbar MRI. Where should it happen?
SYNTHETIC DEMONSTRATION DATARoutine urgency, set by her physician. Demonstration PPO, deductible already met. The MOM evaluates three imaging alternatives in their whole operating context.
- Cost
- $3,200
- Patient responsibility
- $640
- Availability
- 11 days
- Distance
- 6 miles
- Network
- In network
- Clinical requirement
- Satisfied
- Cost
- $780
- Patient responsibility
- $156
- Availability
- 2 days
- Distance
- 9 miles
- Network
- In network
- Clinical requirement
- Satisfied
- Cost
- $610
- Patient responsibility
- $366
- Availability
- 1 day
- Distance
- 4 miles
- Network
- Out of network
- Clinical requirement
- Not satisfied
- Clinical requirements satisfied
- Prior authorization confirmed
- In network
- Earlier appointment: 2 days, not 11
- Required imaging capability available
- Quality requirements satisfied
- Travel acceptable: 9 miles
- Lower patient responsibility: $156, not $640
- Lower total system cost: $780, not $3,200
- Results sent electronically to the ordering physician
A recommendation the Physician, Payer and Patient can see and explain: what was recommended, why, what information influenced it, which tradeoffs were made, what economics were involved and where human judgment remains.
All figures are illustrative demonstration data. Not a clinical recommendation and not a price from any real Payer or Provider.
Price is one variable. Price is not the objective function.
The lowest-priced option in the demonstration, the Open MRI Clinic, is also the soonest and the nearest. It is not recommended: its scanner does not meet the protocol the physician ordered, it is not authorized at that site and it is out of network. The Healthcare MOM evaluates every dimension at once.
When the Operating Model governs the whole, interests can reinforce one another
The objective is not to optimize the Patient, Provider and Payer independently. It is to find operating decisions where their interests reinforce one another. In Jane’s case, one decision benefits everyone involved:
- PatientThe MRI 9 days earlier, and $484 less out of pocket.
- Ordering physicianThe diagnostic information 9 days sooner, sent electronically.
- Payer$2,420 less for the same diagnostic information.
- HospitalScarce scanner time preserved for patients who need hospital-level resources.
- Imaging centerOpen capacity used for work it is equipped to do.
- Healthcare systemThe same required diagnostic information, produced with fewer resources.
Multiple participants benefit because the Operating Model optimized the behavior of the system rather than one component. The point is not that Payer, Provider and Patient interests always align. They do not. The point is that Operating Architecture can find decisions where better system design produces mutually reinforcing benefits, and expose and govern the tradeoff where it cannot. Constructive Interference is BlueHour’s analogy from physics.
MRI equipment is also a major capital asset. A Provider may have millions of dollars invested in imaging capacity, so one imaging decision can affect:
That does not mean maximizing MRI utilization regardless of clinical need; that would be component optimization. In Jane’s case the system decision preserved scarce hospital imaging capacity and made better use of another appropriate imaging asset, in the same decision that got her imaged sooner at lower cost. That is Capital Discipline, MOM-001, applied inside healthcare.
Lowest cost does not necessarily produce the highest-value outcome
Healthcare involves legitimate conflicts. In a second synthetic case, Thomas Reed needs the same MRI, but his physician has documented a change in his condition and marked the order urgent: imaging within 3 days. No lower-cost site has capacity inside that window.
- Total cost
- $700
- Patient pays
- $140
- First appointment
- 14 days
- Meets the physician’s window
- No
- Total cost
- $2,100
- Patient pays
- $420
- First appointment
- Tomorrow
- Meets the physician’s window
- Yes
The MOM recommends the more expensive option. Clinical timing set by the physician governs before cost. The Payer spends $1,400 more and Thomas $280 more, and both see why before the appointment. The MOM does not set urgency. It weighs clinical requirements, time, quality, economics, patient circumstances, capacity, network, authorization and risk together, and Operating Architecture makes the tradeoff visible, explainable and governed.
Humans must be the loop
Nothing here is autonomous medical decision-making. BlueHour does not replace physicians, clinical judgment, patient choice, Payer responsibilities or human accountability. The architecture coordinates intelligence; it does not eliminate human responsibility. BlueHour provides the Operating Architecture for combining AI × IT × Human Intelligence, so each participant’s intelligence becomes part of a governed system.
One MRI is small. Multiply the decision across healthcare.
Small improvements in medical cost, out-of-pocket cost, administrative effort, scheduling, capacity, duplication, time to diagnosis and patient experience compound when the same architecture governs every decision of its kind. Here, if 1 routine imaging decision in 5 has a better-governed site:
| 10,000 DECISIONS | 50,000 DECISIONS | 100,000 DECISIONS | |
|---|---|---|---|
| Decisions with a better-governed site | 2,000 | 10,000 | 20,000 |
| Total medical cost avoided | $1.8M | $9.2M | $18.4M |
| Patient out-of-pocket avoided | $360K | $1.8M | $3.6M |
| Days to diagnosis returned to patients | 8,000 | 40,000 | 80,000 |
| Administrative hours released | 2,500 | 12,500 | 25,000 |
| Duplicate scans avoided | 50 | 250 | 500 |
| Hospital scanner slots preserved | 1,000 | 5,000 | 10,000 |
Across 1,000,000 healthcare decisions of any kind, an improvement of $25 a decision is $25.0M, and 5 minutes saved on each is 83,333 administrative hours.
Small Improvements × Large Operating Volume = Operating Leverage
In a real engagement, a Payer or Provider would replace these assumptions with its own utilization, pricing, capacity, administrative cost, authorization data, patient economics and clinical requirements.
Illustrative assumptions, set deliberately below Jane’s case. Not ROI, not a forecast and not a result from any Payer or Provider. Change the assumptions in the demonstration →
What other healthcare operating problems could we solve this way?
The MRI is the demonstration vehicle, not the definition of the MOM. The underlying operating problem may prove to be Diagnostic Imaging Decisioning, Diagnostic Decisioning, Site-of-Care Decisioning or something else. The MRI demonstrates the architecture. It does not constrain it. BlueHour will use conversations with Payers, Providers and other healthcare executives to find where the most important operating problems and commercial opportunities are, and only then name the MOM.
A possible healthcare tray. These are potential Micro Operating Models, not existing BlueHour products. Each would be activated inside the Operating Model, not installed like an application.
Healthcare is another application of the same architecture
Not a separate company, and not a disconnected healthcare product. Healthcare MOMs would work with the rest of the BlueHour tray:
Patient outcomes and business performance are not opposing objectives
For CEOs, CFOs, CIOs, COOs, clinical leadership and healthcare Boards. A properly engineered Operating Model looks for decisions that improve both.
We will show you what it looks like as a Micro Operating Model.
For healthcare Payers, Providers, health systems, healthcare executives, clinical leadership, CFOs, CIOs, COOs, Boards and strategic partners.
Or email info@bluehourtechnology.com
Healthcare does not lack intelligence. It has enormous amounts of it, distributed across people, organizations, technologies and systems. Getting that intelligence to operate together toward the desired outcome requires architecture. The objective is Positive Patient Outcomes and Experiences and, wherever possible, better clinical results, better patient experience, better economics and better use of healthcare resources together.
In healthcare, the governed outcome is Positive Patient Outcomes and Experiences.
A conceptual enterprise demonstration. It does not provide medical advice, diagnose patients or recommend treatment, and it does not replace licensed medical judgment. All patient, provider, payer, pricing and scheduling information is synthetic.