Seven parts in, the argument holds together, but it rests on an assumption I have quietly deferred until now. You can only manage what you can see.
You cannot pursue value without measuring outcomes. You cannot pay for resolution without tracking whether the case closed. You cannot target the costly few without stratifying the population. You cannot prove any of it worked without a counterfactual. Every one of those is a measurement problem, and it is where most reforms quietly die.
So this final piece is about the plumbing. Less inspiring than the vision, and far more decisive.
The thing nobody measures
Recall Porter's point from Part 5: of the dozens of quality metrics most systems collect, almost none are outcomes. They are process. Did we follow the steps, run the test, hit the target.
The one thing missing is the only thing that matters to the patient: are they better? That is captured by what the field calls patient-reported outcome measures, the direct question to the patient about pain, function, and whether life improved. Bodies like ICHOM have spent over a decade building standard sets of exactly these. The tools exist. Most systems still do not use them at scale.
Why not? Because it is genuinely hard:
- Building, validating and translating outcome measures for each condition is slow and expensive.
- Getting the right questionnaire to the right patient at the right moment, then wiring the answer back into the record, defeats most IT systems.
- Clinicians, already overloaded, resist another data-entry task that seems to serve managers rather than patients.
None of that is a reason to skip it. It is a reason to treat measurement as the core build, not an afterthought.
Data you have versus data you need
Systems drown in data and starve for information. The distinction matters.
Most health data is administrative: billing codes, activity counts, bed-days. Useful for paying invoices, nearly useless for judging value. It tells you what was done and to whom, never whether it helped.
The data that runs a value-based system is different in kind:
- Outcomes, reported by patients, not just clinical proxies.
- Full-cycle cost, tracked across the whole condition, not per line item. Very few systems can actually total the cost of one patient's journey across settings.
- Linked records that follow the patient between primary care, hospital and home, so the case has one continuous story rather than a dozen disconnected fragments.
The last one is the quiet foundation of everything in this series. The patient who owns her own coordination, from Part 2, exists because her record does not travel. Fix the record and you have fixed a large part of the ownership problem by default.
The feedback loop is the point
Here is the mental shift. Measurement is not for the audit. It is for the learning.
Porter's real argument was that without measuring outcomes, providers cannot improve, because they never find out what works. A closed feedback loop, you treat, you measure the result, you adjust, is how any complex system gets better over time. Healthcare, uniquely, has mostly run without one. It has been flying on process compliance and hoping.
Measure outcomes and cost per patient, feed it back to the people delivering care, and improvement stops being a directive and becomes a habit.
That is the whole machine: not a dashboard for the ministry, but a loop that lets a care team see its own results and get better.
How you actually build it
You do not build this by boiling the ocean. The systems that get there tend to sequence it:
- Start with one condition, or one segment. Usually the complex, high-cost group from Part 7, where the value at stake justifies the effort. Prove the loop on a narrow front.
- Measure outcomes and full-cycle cost for that group. Even roughly. Directionally right beats precisely useless.
- Attach the payment to it. Shared savings gated on the outcomes, per Part 6, so the incentive and the measurement point the same way.
- Give one owner the case and the data, per Part 4, so accountability and information sit in the same hands.
- Then widen. Once the loop works for one group, extend it, condition by condition. This is roughly how the credible integrated-care programmes grew, not by decree but by expansion from a working core.
The end of the series
Strip eight parts down to a sentence and it comes to this. European healthcare is not short of medicine or, mostly, of money. It is short of ownership, of the right incentives, and of the ability to see its own results.
None of the fixes are exotic. Own the case. Pay for the outcome. Concentrate on the few who need it most. Measure what actually happens and learn from it. The reason they are rare is not that they are unknown. It is that each one demands giving up something comfortable: the volume that fee-for-service pays for, the control that a public monopoly guards, the busywork that looks like progress.
The diagnosis was always the easy part. The design is the work. And the design is now, finally, within reach.
Thank you for reading.
Sources
- Porter M.E. (2010), "What Is Value in Health Care?", New England Journal of Medicine 363:2477-2481. On outcomes versus process measurement and the feedback loop for improvement.
- International Consortium for Health Outcomes Measurement (ICHOM). ichom.org. On standardised patient-reported outcome sets.
- On PROM implementation difficulty: reviews in BMC Medical Informatics and Decision Making and BMC Health Services Research (2021-2022).
- OECD (2019), Health in the 21st Century: Putting Data to Work for Stronger Health Systems. On the gap between administrative data and information usable for outcomes and value.
Part 8 of the series, and the final piece. Earlier parts covered the foundation, coordination, misrouting, ownership, value, payment, and the economics of the few.
Published by CW1. Malm is CW1's continuous-care programme for patients: malm.care. Nortb is CW1's operations partner for clinics and practices: nortb.com.