Clinical Data Portability
Definition
The cross-clinic, cross-vendor portability of clinical data needed to track a patient’s chronic disease over years. Surfaced by ari-leshno in 2026-05-19-saar-leshno-glaucoma-data-tracking as the structural pain behind the glaucoma visual-field tracking wedge — but explicitly generalizable beyond ophthalmology to any chronic disease where treatment hinges on multi-year trends.
Key points
- Why it’s painful. Chronic diseases (glaucoma, diabetes, hypertension) require multi-year longitudinal data to drive treatment decisions. When a patient moves between clinics or providers, the prior data is often only available as printed PDFs that the new clinician has to visually compare — or is missing entirely, forcing redundant exams or sub-optimal care.
- Why it persists. Instrument vendors (e.g. zeiss in ophthalmology) have no commercial incentive to support cross-vendor import — proprietary export formats are part of the lock-in that drives instrument sales. DICOM is theoretically the standard but isn’t actually interoperable in practice.
- The today-workaround is paper. The literal current workflow at sheba is: instrument → print PDF → re-scan into the kameleon EHR → manually compare PDFs visually. Even at top-tier hospitals.
- Wedge product shape. OCR-based PDF ingest + tracking UI + low-friction subscription pricing (~100K Zeiss Forum) or that need to track patients who arrive with cross-vendor history.
- US context inverts intuition. Per ari-leshno, US EHRs are “a decade behind Israel” — many clinics still on paper and pen — which means the integration baseline is lower than expected. That actually favors a PDF-OCR product over a deep-integration play.
- Generalizable beyond ophthalmology. Diabetes (lab-result trends), hypertension (BP-cuff trends), and other chronic-condition tracking share the structural shape. Whether one product can serve multiple verticals or each needs a vertical-specific build is open.
Evidence
- 2026-05-19-saar-leshno-glaucoma-data-tracking — Ari Leshno’s pitch; concrete sizing for the glaucoma instance (35K clinics × 10% × $100/mo ≈ tens of millions/year); analogy to diabetes and hypertension.
- 2026-05-01-entree-capital-enterprise-ai-spend-map — external timing tailwind: healthcare named as one of three 2026–27 vertical AI breakouts (alongside legal and financial services); first vertical AI IPOs expected 2027–28 (Bessemer). The ari-leshno glaucoma wedge sits inside this window. See also vertical-ai-tam for the 450B SaaS framing that justifies vertical-healthcare as a fundable category.
Open questions
- Does the cross-clinic problem actually generalize across chronic-disease verticals, or is each instrument ecosystem too idiosyncratic to share a product?
- Does any vendor (other than zeiss in ophthalmology) own a “Forum-equivalent” longitudinal-tracking product in another vertical (diabetes, hypertension)? If yes, who?
- Is OCR-from-PDF a sustainable acquisition strategy, or a wedge that gives way to direct integrations once a clinic adopts? Different unit economics.
- Does the Brain / agent layer add meaningful value beyond a CRUD-with-OCR product? If not, this is a healthcare SaaS, not a Brain vertical — see vertical-use-case-led-brain for the test.
- FDA gating: can the SaaS layer ship without regulatory clearance, with the predictive-model layer (idx-style “scan is normal” framings) bolted on later? Probably yes, but not confirmed.