Ask a room of clinicians why a good idea died, and you will hear five different answers before anyone finishes their coffee. The surgeon blames procurement. The CIO blames the clinicians. The finance director blames the pilot that never had a budget line. The nurse on the ward blames the fact that nobody asked her.
They are all describing the same corpse. They just disagree on the cause of death.
So we want to run a simple poll. Not “what are the barriers to innovation”, because the honest answer to that is “all of them”. The question is sharper: in the system you actually work in, which single barrier does the most damage? Pick one. Then tell us why in the comments.
Here are the options.
- A. Payment: the way we pay for care rewards activity, not better outcomes, so nobody is paid to adopt the thing that reduces activity.
- B. Procurement and regulation: approval, tendering and compliance cycles move slower than the technology they are meant to evaluate.
- C. Workforce capacity: staff are too stretched to learn, test or champion anything new.
- D. Data and interoperability: systems cannot share information, so every new tool becomes another island.
- E. Culture and leadership: risk aversion, hierarchy and short leadership tenures mean nobody owns a change long enough to finish it.
Before you vote, here is the case for each. Every one of them is a genuine contender, and the argument you find weakest may be the one your colleagues pick.
A. Payment: nobody gets paid to do less
The strongest version of this argument is uncomfortable because it does not require anyone to behave badly. Under fee-for-service or activity-based funding, a hospital that adopts a remote monitoring tool which keeps heart failure patients out of hospital loses admissions. Its costs may fall, but its revenue falls faster. The innovation is good for the patient, good for the payer and bad for the provider that has to buy it.
This is why so many digital tools end up adopted by insurers or integrated systems that hold the whole budget, rather than by the hospitals that deliver the care. If you work somewhere that pays by the procedure, you will recognise the pattern: the business case for anything preventive has to be written by someone who will not see the savings.
B. Procurement and regulation: the clock is wrong
Health products should be regulated. The counter-argument is not about whether, but about speed and fit. Tender cycles designed for buying beds and gloves are poorly suited to software that updates every few weeks. A tool can be clinically validated, approved and still spend a year or more in local information governance, security review and contracting before a single patient sees it.
Smaller companies often do not survive that wait. Larger incumbents do, which is one reason health systems end up buying from the same handful of suppliers. If your honest diagnosis is that good ideas die in the queue rather than in the evidence, this is your vote.
C. Workforce capacity: no slack, no change
Innovation needs spare capacity. Someone has to learn the new pathway, run it in parallel with the old one, spot what breaks and fix it. In systems where rotas are filled with agency staff and every clinic is overbooked, that slack does not exist.
The telling sign is who ends up doing the implementation work. If it is a handful of enthusiasts doing it on their days off, the project will last exactly as long as their stamina. Many “failed” innovations did not fail on merit. They failed because the one person carrying them burned out or moved job.
D. Data and interoperability: every tool becomes an island
A decision-support tool is only as good as the data that reaches it. When lab results live in one system, prescriptions in another and community care notes on paper, each new product has to build its own integrations or ask staff to type the same information twice.
The cost here is cumulative. The tenth good tool is harder to adopt than the first, because each one adds another login, another interface, another data feed that breaks when a supplier updates something. Clinicians start rejecting new tools not because they doubt the tool but because they doubt the plumbing.
E. Culture and leadership: nobody stays long enough
Healthcare is right to be cautious. The problem is when caution becomes the default answer to everything, and when the people with authority to say yes change every two or three years.
Innovation in health systems usually takes longer than a single leadership tenure. A chief executive or minister who sponsors a programme often leaves before it matures. The successor inherits a half-finished project with someone else’s name on it, and a strong incentive to start something new. Add a hierarchy in which junior staff who see problems first have the least power to fix them, and you get organisations full of ideas that nobody is allowed to finish.
Why the vote matters
These barriers interact, which is exactly why the answer is worth arguing about. Payment reform does little if procurement takes two years. Better data does little if nobody has time to use it. But systems cannot fix everything at once, and the order matters. A health ministry that believes its main problem is culture will invest in leadership programmes. One that believes it is payment will redesign tariffs. They cannot both be right about where to start.
Your answer probably depends on where you sit. We suspect clinicians and managers will vote differently, and that people in publicly funded systems will vote differently from those in insurance-based ones. That split is the interesting part.
The strongest objection to each option
Every option above has a good rebuttal, and you should hear it before you vote.
Against payment: plenty of systems that hold the whole budget, where the savings from prevention land in the same pocket that paid for it, still fail to adopt good tools. Integrated systems and national health services are not visibly more innovative than fee-for-service ones in every area. If aligned payment were sufficient, those systems would be running away from the pack. They are not, which suggests payment is necessary but not decisive.
Against procurement and regulation: slow approval is often a symptom rather than a cause. Information governance teams take months because nobody has given them the capacity, the standard contracts or the clear risk appetite to do it faster. Blaming the queue can be a way of avoiding the harder question of why the organisation never resourced the people who run it.
Against workforce: stretched staff do adopt new things when those things save them time immediately. Many clinicians took up remote consultations almost overnight during the pandemic, under more pressure than ever. So exhaustion does not block every change. It blocks changes that cost time before they return it.
Against data: interoperability is a real problem, but some of the most successful innovations in health care needed no integration at all. A checklist, a changed rota or a new referral rule can travel through any IT estate. If your system never adopts even those, the plumbing is not the whole story.
Against culture: “culture” is where diagnoses go when nobody wants to name a specific mechanism. It is true, it is unfalsifiable and it tends to generate leadership programmes rather than decisions. A vote for E is strongest if you can point to the specific behaviour, such as who is allowed to say no and how often, that you would change.
A quick diagnostic for your own system
If you are torn, try tracing the last good idea you saw die and ask where it stopped.
If it died because the business case showed a loss for the organisation that had to buy it, even though the patient and the wider system would have gained, that is payment.
If it had evidence, a willing clinical team and a budget, then spent most of a year in contracting, security review and governance before being quietly dropped, that is procurement.
If it launched, worked in the pilot ward and then faded when the champion went on leave or changed job, that is workforce.
If staff liked it in principle but stopped using it because it meant typing things twice, logging into another system or reconciling data by hand, that is interoperability.
If it never got a clear yes or no, bounced between committees, and lost its sponsor at the next reorganisation, that is culture and leadership.
Most real cases involve two or three of these. The one you should vote for is the first place it stalled, because fixing the later barriers does nothing if the idea never gets that far.
How the barriers feed each other
The options are listed separately, but in practice they form loops.
Weak payment incentives mean no one funds the implementation team, which makes workforce capacity the visible failure. Poor interoperability means every new tool needs a bespoke integration, which lengthens procurement and security review. Short leadership tenure means nobody sponsors the multi-year work of fixing data standards, so interoperability never improves. A system can spend a decade treating symptoms of one barrier while the root sits in another.
This is also why the “pilot graveyard” is so full. Pilots are usually funded from a special pot, staffed by enthusiasts and exempted from normal procurement. They succeed precisely because they bypass the barriers. Then scale-up requires the idea to survive the ordinary budget, the ordinary workforce and the ordinary buying process, and it dies. A pilot that succeeded under exceptional conditions has proved very little about whether it can survive normal ones.
What each answer implies for the people in charge
The vote is not academic, because each answer points to a different owner and a different first move.
If payment wins, the responsibility sits with payers and ministries: tariffs that pay for outcomes or for the whole pathway, shared-savings arrangements, and a willingness to fund one organisation so that another saves money.
If procurement wins, the fix is operational and surprisingly cheap: standard contract templates, shared assurance so that a product checked by one organisation does not have to be rechecked from scratch by the next, and teams sized for the volume of work.
If workforce wins, implementation time has to become a budgeted cost of any new product, protected in job plans, rather than something clinicians do in the margins.
If data wins, the lever is standards and enforcement: requiring suppliers to use open interfaces as a condition of sale, and refusing to buy tools that cannot connect.
If culture wins, the change is about decision rights: who can approve a small, reversible trial without a committee, and how long a sponsor stays accountable for a programme after they move on.
What we will be watching
When the comments come in, we will look for patterns rather than a single winner. Do clinicians blame capacity while managers blame procurement? Do people in insurance-based systems pick payment more often than those in tax-funded ones? Does anyone make a convincing case that we missed the real barrier entirely? We will write up what we find.
How to vote
Reply in the comments with your letter (A to E), your role and the kind of system you work in. Then give us one concrete example: a good idea you watched die, and what killed it. If you think we have missed the real barrier, make the case for option F.
We will not pretend to have the answer. Tell us yours.
