The Geneva Protocol: Design Notes
Why our 30-day initiation looks the way it does — and the four things we changed after the first 400 patients told us we were wrong.
The Geneva Protocol is the name we use internally for the 30-day initiation that every BSerum patient passes through before entering maintenance. It is the most heavily revised document in our clinical playbook — currently on version 4.2 — and almost every revision was driven by something a patient told us during a session debrief.
1. The shape of the 30 days
- Day 0: Full biomarker baseline, IgE and basophil panel, body composition scan
- Days 1-3: Three micro-priming doses at 0.1× target
- Days 4-10: Three sessions at 0.4× target with full Guardian telemetry
- Days 11-21: Three sessions at 0.7× target, biomarker reassessment at day 14
- Days 22-30: Two sessions at full target, exit panel, maintenance plan
2. What we changed in v4.0
Originally we ran a five-priming-dose ramp over the first week. Patient diaries consistently showed that the third dose was the one that mattered for tolerance; doses four and five added nothing measurable and asked patients to come into the clinic twice more than was clinically useful. We compressed the priming arm to three doses and watched our event rate stay flat at zero. We also gave patients back two days.
3. What we changed in v4.1
We moved the day-14 biomarker reassessment from a clinic visit to a home capillary draw with same-day courier. Adherence to the reassessment went from 71% to 98%. The dataset got better, the patients were happier, and our cost per session fell.
4. What we changed in v4.2
We let the Guardian halve the time between micro-doses inside a single session if patient-reported pain stays below threshold for two consecutive cycles. This means sessions can finish 18 minutes earlier on average without changing the total delivered dose. The protocol is dynamic now, not fixed.
| Version | Released | Headline change | Outcome |
|---|---|---|---|
| v3.0 | 2024-Q3 | Initial 30-day ramp | Baseline event rate: 0 |
| v4.0 | 2025-Q1 | Compressed priming 5→3 | −2 visits per patient |
| v4.1 | 2025-Q2 | Home capillary draw at day 14 | Adherence 71%→98% |
| v4.2 | 2025-Q4 | Adaptive in-session pacing | −18 min mean session length |
"Every version of the protocol exists because a patient told us, in plain language, what was wrong with the version before it."