The 14 synthesis tools

Each market-access tool sits on the same workspace shell with a six-tab strip (Sources · Review · Evidence · Build · Export · Audit); its distinctive surface is the Build tab. Every tool is agency-aware — it adapts currency, thresholds, required scenarios, and sections to the launch's HTA agency.

Scoping

• PICO scope — an editable Population / Intervention / Comparator / Outcomes form with an agency-aware validation panel (preferred outcome, required sections, time-horizon defaults, comparator-framing rules) and a matching-analogues sidecar.
• Comparator & analogue navigator — filter chips over a table of analogue decisions (PSDs, TAs, IQWiG appraisals) with fit-score bars; each row shows comparator framing, committee notes, ICER, and a "use as CEA comparator" action.

Economic core

• Population estimator — a five-year eligible-patient forecast with a confidence interval and four scenarios (base, optimistic, pessimistic, workforce-constrained). Drivers carry attribution chips; saved runs carry a reviewer signature and feed the budget-impact and cost-effectiveness tools.
• Budget impact (BIM) — reads the estimator run; a headline year-5 net-impact KPI, a five-year stacked-bar chart (drug cost vs offsets vs net), the agency's required scenarios, and driver-elasticity cards.
• Cost-effectiveness (CEA) — a model-structure picker (Markov, partitioned-survival, decision tree), a transition matrix, utility weights, ICER plus delta-cost / delta-effect / net monetary benefit / probability cost-effective, a cost-effectiveness-plane scatter, and an agency-specific required-sections checklist. Switches to efficiency-frontier mode for IQWiG.
• Sensitivity & threshold — three modes: tornado (top inputs by elasticity), threshold (the price that hits the willingness-to-pay), and PSA (a 1,000-point scatter and CEAC).
• Pricing & contracting — ex-factory and net price across markets, a reference-pricing scenario grid, and a price-sensitivity readout.

Readiness

• Evidence gap — a submission-readiness percentage and a matrix of required parameter by consumed-by-tool by confidence by status; one-click Source and Review buttons inline.
• Equity impact — agency-aware distributional dimensions across metro / regional / remote bands (and agency-specific dimensions such as IMD decile or indigenous status).

Strategy

• Submission timeline — a Gantt across nine milestones over the HTA cycle calendar with critical-path dependencies and target / backup windows.
• Resubmission strategist — clusters the top failure modes from prior outcomes (population over-estimation, comparator framing, ICER uncertainty, cohort representativeness) and drafts a recovery plan. Active when a launch is rejected.
• Committee Q&A — likely committee questions with a risk score and cited draft answers.
• Committee simulation — a probability distribution over verdicts, a ranked question list, and a risk register.
• Cross-HTA translator — a field-by-field diff from a source-agency dossier to a target, tagging each field as carries (no change) or remap (needs a target value).
• KOL & evidence network — a graph of key opinion leaders, publications, and committee relationships per indication, with influence scores and engagement status.

The Wizard

The Wizard (sidebar hero, or ⌘K) renders the matching tool widget inline first, then asks. It classifies intent from the prompt, cites every source, and refuses ungrounded questions. Every mutation requires an explicit confirm and appends an audit row. See the screver-ma-ai service for the grounding implementation.