Making the Rare Reliable

For conditions that fall below what regulators call ‘very rare’ – roughly 1 in 50,000 people or fewer – the standard evidence machinery doesn’t just run slow. It often runs out before patients can afford to wait. The National Institute for Health and Care Excellence (NICE) routes technologies for conditions below that prevalence threshold to its Highly Specialised Technologies programme precisely because they sit outside the territory where normal appraisal assumptions hold; the Scottish Medicines Consortium draws the same line for ultra-orphan medicines. When the UK’s Medicines and Healthcare products Regulatory Agency (MHRA) opened a consultation on a new rare disease therapies framework, it acknowledged that traditional development programmes for rare conditions typically take 10–12 years to reach marketing authorisation – a timeline that, for patients with rapidly progressing conditions, is functionally indistinguishable from no treatment at all.

Regulators have started to say this out loud. In draft guidance on frameworks for individualised therapies in ultra-rare diseases, the FDA anticipates situations where randomised controlled trials are not feasible because patient populations are too small. When volume can’t generate the certainty, structured codification of expert judgment – explicit decision rules, defined verification steps, auditable protocols – is what stands in its place.

Rarity Demands Rigour – A Cross-Domain Pattern

Across high-consequence fields, the recurring answer to low-frequency, high-stakes problems is not to wait for more experience; it is to codify the judgment of those who have already faced them. When events are too rare for most practitioners to build reliable intuition, the alternative is to capture expert decision logic in structured, auditable procedures so that knowledge gathered through uncommon exposure becomes available to those encountering the scenario for the first time. This is not a workaround or an optional extra. It is the rational response to rarity when improvisation carries costs that cannot be absorbed as learning experiences.

Aviation, nuclear operations, and structural engineering all formalise this logic. Federal Aviation Administration guidance frames standard operating procedures as covering normal, abnormal, and emergency operations; International Atomic Energy Agency safety guides for nuclear power plants include emergency operating procedures and severe accident management guidelines; and US Department of Defense criteria for progressive collapse mitigation describe that type of collapse as a relatively rare event yet prescribe standardised design methods to limit it. Across these domains, rare but severe scenarios are explicitly written into procedures because people managing them cannot rely on routine exposure to keep their skills calibrated.

The current European work on small modular reactors applies the same principle in regulatory form. In January 2026, France’s nuclear safety and radiation protection regulator ASNR announced the launch of the third phase of a joint review of NUWARD small modular reactor safety options, with ASNR and seven counterpart regulators jointly examining how safety challenges are handled across national frameworks. As ASNR stated, “The conclusions of this review, expected by the end of 2026, will inform ASNR’s work on harmonising safety requirements and authorisation processes for new reactors.” Harmonisation here means agreeing the decision rules and verification expectations in advance for operational scenarios that will be rare, novel, and high-consequence. Medicine faces the same structural conditions – rare, high-stakes decisions where intuition built on routine exposure is unavailable – and the same design logic applies.

What Genuine Codification Looks Like

There is a crucial difference between a genuine clinical pathway and a document that simply describes what a particular clinician tends to do. Genuine pathways specify indications, required equipment, role allocations, and verification steps, and define what should trigger reassessment or deviation. Preference documents are organised around a person rather than a protocol and cannot be meaningfully audited, transferred, or improved. Implementation research on surgical safety checklists shows how easily codification slides into this weaker form. In a qualitative study of checklist use, health services and implementation researcher Anna R. Gagliardi observed that “Extensive modification to accommodate existing practice patterns eliminated essential interaction at key time points to discuss patient management.” When local adaptation preserves the paper tool but removes the interaction moments that create shared situational awareness, the protocol’s functional mechanism has already failed – even though the checklist appears present and complete.

A province-wide evaluation in Ontario made the same point at system scale. After surgical safety checklists were mandated across hospitals, an observational study comparing outcomes before and after implementation found the checklists were not associated with statistically significant reductions in operative mortality or major complications. In that setting, a mandatory document and recorded completion did not, by itself, produce measurable improvements in patient outcomes. The lesson is not that checklists are ineffective; it’s that codification confined to paperwork is often indistinguishable from no codification at all.

Atlantoaxial osteoarthritis – degenerative disease at the C1–C2 junction of the cervical spine – sits squarely in the category of rare, high-stakes problems where volume-dependent evidence is thin and ordinary experience is scarce. For this problem, the relevant quality lever is not another randomised trial but the design of a pathway that makes a complex, risky reconstruction reliably executable. Dr Timothy Steel, a neurosurgeon and minimally invasive spine surgeon practising at St Vincent’s Private and St Vincent’s Public Hospitals, has established a complex cervical reconstruction pathway for atlantoaxial osteoarthritis built around image-guided posterior C1–C2 fixation. The pathway standardises preoperative CT and MRI planning, the use of Brainlab stereotactic navigation in theatre, posterior fixation using transarticular screws and Harms constructs, and defined postoperative imaging to confirm fusion, with theatre staff trained on the specific navigation and fixation systems the protocol requires. Naming the equipment and the training requirement explicitly is not incidental detail – it’s what separates a pathway another team could enact from a narrative only its author could reproduce.

An external study of 23 patients treated between 2005 and 2015 under this approach reported Visual Analogue Scale pain scores falling from 9.4 to 2.9 and 95.5% radiographic fusion. Those outcomes were reported under a defined, auditable pathway – not under undocumented individual judgment – which is the relevant distinction when reliability and transferability are the measure.

The Infrastructure Behind the Pathway

Well-designed pathways do not maintain themselves. The conditions that make a complex, low-volume protocol executable – trained staff, calibrated equipment, rehearsal of rare workflows, and referral patterns that bring the right patients to the right environment – are easy to miss because they don’t appear on simple volume or throughput reports. Counting how many procedures a centre performs says little about whether it has protected the team familiarity, configuration discipline, and planning capacity that rare, high-stakes work depends on.

A January 2026 systematic review in Annals of Surgery on centralised complex surgery makes this visible in outcome data. Reviewing multihospital systems that concentrated complex surgical care in high-volume centres, the authors reported relative reductions in perioperative mortality of roughly 15–40%, 10–30% reductions in major morbidity, and hospital stays shorter by about 0.5–2 days. Crucially, the benefits were strongest where centralisation went hand in hand with formal referral pathways, standardised perioperative protocols, and continuous outcomes monitoring.

Those findings suggest that better outcomes are not a simple function of doing more cases; they arise when higher volume is coupled to the organisational infrastructure that makes pathway-based care possible – predictable referral routes, consistent perioperative routines, clear team roles, and feedback loops that detect drift. Where that infrastructure is absent, shifting cases alone cannot reliably generate the same gains, because the conditions that pathways depend on have not been built.

When the Conditions Disappear

Pathways degrade quietly. The knowledge embedded in a well-designed protocol is distributed across people, tools, and institutional routines – clinicians who understand indications and edge cases, teams rehearsed in the critical sequence, equipment and implants that match what the protocol specifies. When any of those elements changes – a key clinician moves on, equipment is upgraded out of spec, experienced staff are rotated elsewhere – the pathway can erode while its documentation sits unchanged, still clearing every governance audit. The same evidence base that supports concentrating complex care in high-volume centres also cautions that weakly organised centralisation can do the reverse – widening access disparities for patients who cannot easily travel to hub facilities.

Because standard dashboards focus on counts and averages, this kind of erosion can remain invisible until a patient encounters the resulting gap: the operation is still nominally “available” and the protocol still exists in the policy manual, but the institution no longer has the conditions to deliver it with the reliability that justified its adoption. NICE’s rare disease quality standard, published in February 2026, makes these dependencies structural obligations rather than aspirational targets – calling for named healthcare professionals, defined diagnostic and treatment pathways, equitable access to recommended therapies, and routine data collection and monitoring. Recognising those dependencies shifts the question from whether to codify expertise to how to steward what has been codified, so that pathway reliability is protected as staff, technology, and organisational structures evolve.

Stewardship of What Has Been Built

The regulatory moves now in motion – rare-disease frameworks from the MHRA, draft FDA guidance for ultra-rare individualised therapies – accept an uncomfortable premise: for very small patient populations, medicine’s standard evidence infrastructure will often arrive late, if at all. That’s not a temporary gap awaiting a methodological fix. It’s a structural feature of conditions where trial-ready populations don’t exist and may never.

Within healthcare, national rare-disease quality standards and the evidence base around centralising complex surgery both point in the same direction. Formal referral routes, defined diagnostic and treatment pathways, standardised perioperative protocols, and routine outcomes monitoring make care for rare and complex conditions more predictable and less dependent on where a patient happens to present – but the same review warns that poorly designed centralisation can deepen access disparities. The architecture of codification matters as much as its presence.

The atlantoaxial osteoarthritis pathway described earlier makes the abstractions concrete: specified indications, imaging, navigation, fixation constructs, staff training, and postoperative verification, assembled into a form a prepared team can enact and an institution can audit. Protecting that value is not a matter of maintaining paperwork but of maintaining capability – trained teams, compatible equipment, organisational memory, and deliberate succession planning for when personnel and technology change. For ultra-rare, high-stakes conditions, a pathway without that stewardship is not a durable asset. It’s a record of what was once possible.