Bondtrials

Bond Health (Bond Trials) — A closed-loop trial-enrollment dataset that links chart-derived eligibility to paid-media exposure and to the realised clinical outcome. For every patient the record carries: an EHR-derived eligibility match against protocol criteria; whether the patient was reached by paid advertising (Meta and Google campaigns, including which campaign creative and channel); whether they responded, were called and texted, and how many follow-up attempts it took; the pre-screen result against criteria answerable from chart review; the visit booking; informed-consent stage; screen-…

Bondtrials offers (Alternative, Sentiment, Reference) — Not published on this page. Structural estimate: each enrolled patient generates the full funnel chain — EHR match, ad exposure, multiple call and SMS follow-up attempts, pre-screen, consent, randomisation, plus post-enrolment engagement — so order 10^2-10^3 events per enrolled patient. At the volume implied by a per-screened-patient fee model across its site base, order 10^5-10^6 funnel events is plausible; the existence of a 2026 benchmark report implies enough cumulative data to publish aggregate statistics. Confirm cumulative screened-patient count directly..

Patient recruitment response-propensity and contactability modelling (which channel, creative, message and follow-up cadence converts an eligible patient into a screened patient — data almost nobody holds because marketing exposure is rarely joined to clinical outcomes); training and validating automated eligibility-matching models against real screen-failure ground truth; screen-failure reduction and protocol criteria refinement (which criteria are unanswerable from chart review and therefore generate failures); enrolment timeline forecasting and site-capability scoring; trial site finance and cost-per-enrolment modelling from measured staff hours; paid-media effectiveness studies for healthcare recruitment across Meta and Google; patient consent-approach research and trial-access equity analysis (who gets approached and who agrees).

The data is with 2 years of history.