Adaptive Designs (bioSkills)

A Claude Code skill for planning adaptive clinical trials — interim analyses, sample-size re-estimation, arm selection, population enrichment — with the boundary maths done in rpact/gsDesign and the design choices anchored to current regulatory guidance.


Type	Claude Skill
Supplier	GPTomics bioSkills (community OSS, MIT)
Availability	GA — part of the bioSkills collection
Pricing	Free / OSS (MIT) — `rpact`, `gsDesign`, `gsDesign2`, `adaptr`, `simtrial`, `BOIN`, `dfcrm`, `trialr` and `escalation` are separately installed OSS R packages. The commercial alternatives the skill names (FACTS, East/EastHorizon, ADDPLAN) are licensed products and are not installed.
Capabilities	Read/Write — Claude runs the skill’s workflow locally (R), not as an MCP tool
Verified	works · 2026-08-03
Security	cleared · 2026-08-03 — GPTomics/bioSkills MIT confirmed, provenance matches, no advisories

How to install

bioSkills is not an npm package — skills are plain markdown/code read directly by the agent. Clone the repo, then either run the installer for the whole category or copy the single skill directory.

Claude Code — clone and install via the bundled script:
```
git clone https://github.com/GPTomics/bioSkills
cd bioSkills
./install-claude.sh --categories "clinical-biostatistics"
```
The installer copies matching skills into ~/.claude/skills/ (default target). Use ./install-claude.sh --list to preview the skills first.
Claude Code / other agents — copy just this one skill:
```
cp -r bioSkills/clinical-biostatistics/adaptive-designs ~/.claude/skills/
```
(run from inside your clone — the previous step left you in bioSkills/; otherwise replace bioSkills/ with the absolute path of your clone). Install the R packages on first use, e.g. install.packages(c("rpact", "gsDesign", "gsDesign2")).

What it does

Walks a five-stage design process and picks the machinery to match:

Design selection — group-sequential (O’Brien–Fleming, Pocock, Lan–DeMets alpha spending), blinded sample-size re-estimation (Friede–Kieser), unblinded SSR (Cui–Hung–Wang weighted statistics, Mehta–Pocock promising zone), combination tests, seamless Phase 2/3 with treatment-arm selection, adaptive population enrichment, and response-adaptive randomisation.
Pre-specification — encoding boundaries, spending functions, interim timing, the SSR rule and the IDMC firewall in the Statistical Analysis Plan before any data are seen.
Interim analysis — computing the test statistic, conditional power or interim effect estimate for IDMC review of unblinded data.
Adaptation decision — applying the pre-specified rule: stop for efficacy or futility, increase n, drop or select arms, restrict the population.
Final analysis — combining stages with the pre-specified weights or combination test, and reporting bias-corrected estimates where the adaptation induces bias.

Software the skill drives: rpact 4.2+ (primary — group-sequential, SSR, combination tests, enrichment), gsDesign 3.6+ and gsDesign2 1.1+ (boundaries and spending), adaptr (enrichment, RAR), simtrial (simulated operating characteristics), and for dose-finding BOIN (FDA Fit-for-Purpose designation, December 2021), dfcrm (CRM), trialr (EffTox) and escalation.

Primary use cases: sizing a trial with interim looks, choosing a sample-size re-estimation rule, planning a master or platform protocol.

Notes

Regulatory anchors cited in the skill: the FDA 2019 Final Adaptive Designs Guidance (Federal Register 2019-25986, 2 December 2019), the FDA 2022 Final Master Protocols Guidance (March 2022), the ICH E20 draft at Step 2b (25 June 2025) — which the skill is explicit is not final — and an FDA CDER Bayesian methodology draft (January 2026). Treat the draft guidances as drafts when writing an SAP.

Distributed as a SKILL.md (plus reference material) in the bioSkills collection — Claude executes the workflow locally rather than as an MCP server. The upstream skill front-matter name is bio-clinical-biostatistics-adaptive-designs; if invoked as a namespaced plugin command it resolves under the bioSkills plugin, not as a bare /adaptive-designs. Upstream directory: clinical-biostatistics/adaptive-designs.

Complements Trial Reporting (analysis and CONSORT-conformant write-up) and Clinical Trial Protocol (protocol document generation). Sample-size work for fixed designs and the missing-data handling that follows an interim are covered by sibling skills in the same category.

Sources

Installed this tool?

Share feedback — install path, OS, errors, workarounds. The form opens with this tool pre-selected and a link back to this page.