NextConsensus Medical-transition forecasting
Discuss a Program

Resource

The 12-rung ladder: evidence → recognition → authority → adoption → consequence

NextConsensus models medical change as a sequence of recognition states. We observe upstream, forecast resolvable transitions, enable downstream decisions.

Validation Standards Example Transition Brief
Rung 1 upstream

Evidence emergence

New trial results, safety signals, subgroup analyses, novel mechanisms, or new indications enter the public record.

Observable signals:
  • Registrational trial publication
  • Preclinical proof-of-concept
  • Post-hoc subgroup analysis
  • Safety signal in FAERS/EudraVigilance
  • New mechanism of action disclosed

Role: Input to model — the raw material of change

Rung 2 upstream

Replication & strengthening

Independent confirmation, effect consistency across populations, safety coherence, and methodological rigor.

Observable signals:
  • Independent confirmatory trial
  • Meta-analysis with consistent direction
  • Subgroup replication across studies
  • Dose-response relationship confirmed
  • Safety profile coherent with mechanism

Role: Filters noise — distinguishes signal from single-study artifacts

Rung 3 upstream

Expert recognition & KOL convergence

Field-level interpretation shifts: review articles, conference narratives, advisory-board consensus, editorial positions.

Observable signals:
  • Major review article endorsement
  • Conference plenary discussion
  • KOL commentary in high-impact journals
  • Guideline committee pre-meeting materials
  • Specialty society position statement

Role: Social validation — the field converges on interpretation

Rung 4 upstream

Real-world confirmation

RWE accumulation, registry data, claims analyses, practice-pattern shifts, and post-market surveillance.

Observable signals:
  • Registry outcome data published
  • Claims database effectiveness study
  • Comparative effectiveness in routine care
  • Long-term safety in broad populations
  • Practice adoption metrics (NBRx, TRx)

Role: External validity — does the evidence hold outside trials?

Rung 5 resolvable

Guideline action

Formal recommendation change by a recognized guideline body — grade shift, new/retracted endorsement, population expansion/restriction.

Observable signals:
  • Guideline update announced
  • Evidence review commissioned
  • Draft recommendation circulated
  • Public comment period opened
  • Final guideline published

Role: Primary forecast target — public, dated, adjudicable

Example propositions (rungs 5–7):
  • Will KDIGO upgrade SGLT2i recommendation from 2B to 1A by 2027?
  • Will ESC narrow ICD indication for non-ischemic CM by Q3 2026?
  • Will NCCN add PARPi as category 1 for HR+/HER2- by 2027?
Rung 6 resolvable

Regulatory action

Label/indication change, approval action, REMS modification, or withdrawal by FDA, EMA, PMDA, Health Canada, etc.

Observable signals:
  • Supplemental application submitted
  • Advisory committee scheduled
  • Complete response letter issued
  • Label negotiation letters public
  • Approval/withdrawal announcement

Role: Primary forecast target — public, dated, adjudicable

Example propositions (rungs 5–7):
  • Will FDA expand [drug] label to include [population] by Q4 2026?
  • Will EMA withdraw [drug] for [indication] following confirmatory trial failure by 2027?
  • Will FDA add boxed warning for [risk] to [drug class] by mid-2026?
Rung 7 resolvable

Payer/coverage action

NCD, LCD, formulary change, step-therapy revision, or access expansion/restriction by CMS, MACs, or major commercial payers.

Observable signals:
  • NCD reconsideration opened
  • MedCAC meeting scheduled
  • Draft LCD posted for comment
  • P&T committee agenda includes review
  • Formulary decision announced

Role: Primary forecast target — public, dated, adjudicable

Example propositions (rungs 5–7):
  • Will CMS issue NCD expanding [therapy] coverage by Q1 2027?
  • Will major PBM add [drug] to exclusion list for 2027 formulary?
  • Will MAC issue LCD restricting [diagnostic] by end of 2026?
Rung 8 downstream

Institutional policy adoption

Hospital system pathway updates, P&T formulary decisions, clinical protocol revisions, EHR order-set changes.

Observable signals:

Role: Enterprise application layer — customers apply forecasts here

Rung 9 downstream

Clinical adoption

Prescribing behavior shifts, guideline-concordant care metrics, specialist vs. PCP uptake divergence, academic vs. community practice gaps.

Observable signals:

Role: Enterprise application layer — customers apply forecasts here

Rung 10 downstream

Commercial strategy adjustment

Promotional claim changes, field force redeployment, patient support program modifications, pricing/contracting strategy shifts.

Observable signals:

Role: Enterprise application layer — customers apply forecasts here

Rung 11 downstream

Market repricing

Analyst consensus revisions, valuation multiple adjustments, competitive positioning shifts, M&A probability changes.

Observable signals:

Role: Enterprise application layer — customers apply forecasts here

Rung 12 downstream

Patient-access / outcome consequences

Time-to-treatment changes, health equity impacts, long-term outcome shifts, budget impact on health systems.

Observable signals:

Role: Enterprise application layer — customers apply forecasts here

Product architecture

Two products, different evaluation criteria

The ladder architecture makes the product separation explicit. Recognition Intelligence operates at rungs 1–4. Authority-Transition Forecasting operates at rungs 5–7. They have different targets, different evaluation metrics, and different maturity.

Recognition Intelligence

Target: Rungs 1–4

Question: Is an emerging evidentiary pattern likely to become broadly recognized?

Targets:
  • KOL convergence
  • Review-literature consensus
  • Conference narrative change
  • RWE confirmation
  • Field-level interpretation shifts
Value: Greater lead time; strategic intelligence; early hypothesis formation

Challenge: Outcomes harder to define; expert samples unstable; 'consensus' can be socially constructed; resolution often requires judgment

Status: Signal detection / recognition modeling — not yet formal forecasting product

Authority-Transition Forecasting

Target: Rungs 5–7

Question: Is a named authority likely to take a defined public action by a deadline?

Targets:
  • Guideline recommendation changes
  • Regulatory decisions
  • Payer policy changes
Value: More directly actionable; easier to score; cleaner cohorts; more defensible performance claims

Challenge: Later signal; requires upstream states to have already progressed; procedural delays can obscure

Status: First formal forecasting product

Non-events have structure

Stalled-transition taxonomy

A binary outcome (occurred/did not occur) throws away most of the useful causal information. The richer representation identifies where the transition stalled. Every forecast record includes: target outcome, highest attained rung, subsequent status.

Evidentiary failure

Stalls at 1–2

The signal did not replicate or weakened materially. Effect size shrank, safety signal disproven, subgroup not confirmed.

Recognition failure

Stalls at 3

Evidence persisted, but expert interpretation did not converge. Field remains divided, review articles conflict, no narrative coherence.

Translation failure

Stalls at 4→5

Experts converged, but the relevant authority did not initiate or advance action. Evidence-recognition gap not bridged to procedural signal.

Procedural delay

Stalls at 5–7

The transition appeared directionally supported but missed the forecast horizon because of process timing (review cycles, committee schedules, administrative backlog).

Institutional resistance

Stalls at 5–7

The authority declined to move despite strong upstream evidence or recognition. Cost concerns, precedent, political pressure, or competing priorities.

Competing transition

Stalls at 5–7

A different action occurred than the one forecast — e.g., narrowing instead of expansion, restriction instead of removal, conditional instead of full approval.

Ambiguous resolution

Stalls at 5–7

The authority changed language, but not enough to meet the pre-registered threshold. Guideline 'softened' but didn't change grade; label added wording but not new indication.

Late occurrence

Stalls at 5–7

The predicted (implication-ok) event occurred after the specified deadline. Still a forecast miss for the original proposition, but useful for timing-model analysis.

Canonical proposition

What every proposition declares

Target rung
5 — Guideline action
Proposition
Will Society X recommend Therapy Y for Population Z by 2027-12-31?
Upstream states observed
Trial evidence, replication, expert commentary, conference discourse, RWE
Deadline
2027-12-31
Resolution source
Independent resolver (guideline publication)
Forecast threshold
70% (example)
Primary comparator
First formal guideline procedural signal
Secondary comparator
KOL convergence threshold
Primary measure
Brier score
Lead-time measure
Days between NC crossing 70% and final recommendation
Operational constraint
Precision at 70% threshold ≥ predefined minimum

Core metric

Lead time over rung

The core commercial metric is not lead time alone — it is useful lead time at a specified precision, calibration level, and alert burden. 

Lead time = date of target transition − date NC first crossed frozen threshold

Example:
Guideline update:       October 15
NC crossed 70%:         May 3
Lead time:              165 days · Illustrative

Always reported with: precision at threshold, recall, calibration, false-alert burden

Comparator timestamps require explicit rules

KOL convergence

First date when ≥3 independent KOL sources express directional consensus in peer-reviewed commentary or major conference proceedings

RWE confirmation

First qualifying RWE publication meeting prespecified methodological standard (registry size, follow-up, confounding control)

Procedural signal

Announced evidence review, draft recommendation, committee agenda inclusion, or MedCAC meeting scheduled

Final action

Final guideline publication, regulatory approval letter, or NCD/LCD final decision memo

Transition-hazard modeling

The system estimates rung-to-rung transition hazards conditioned on:
- Current evidence state at each rung
- Time already spent at each rung (dwell time)
- Authority-specific behavior (historical pattern)
- Event class (guideline vs regulatory vs payer)
- Disease area
- Precedent (similar transitions in same class)
- Procedural status (review announced, draft published, etc.)

Outputs:
1. Transition Forecast — probability of reaching target rung by deadline
2. Bottleneck Diagnosis — which rung-to-rung transition is currently limiting downstream movement

Example bottleneck output:
Evidence strength: high
Expert convergence: moderate
RWE confirmation: low
Procedural readiness: low
Primary bottleneck: Insufficient real-world confirmation

See a concrete forecast specification.

Inspect the exact format of a registered forecast: proposition, authority, deadline, evidence cutoff, baselines, and resolution rule.

Transition Brief Validation Standards

Registered and scored forecasting for defined medical transitions.