HugoScore hugoscore.org

Full review

TrialScreen CAIHL draft report

Evidence-linked HugoScore draft report for a health AI tool that affects patients.

Back to profile Markdown source

HugoScore CAIHL Draft Report: TrialScreen

Status: Draft for human review Last reviewed: 2026-06-10 Review method: Public-source review of trialscreen.org and info.trialscreen.org homepage, about, researchers, and FAQ pages, privacy policy, Vic Trials partner listing, and Opyl/Opin divestiture press coverage; no platform walkthrough, vendor interview, or independent evaluation of matching accuracy. Service: TrialScreen Vendor: TrialScreen Pty Ltd (with TrialScreen Inc. in the US) Category: Clinical trial matching AI

Summary

TrialScreen is a free global clinical trial search engine at trialscreen.org, operated by TrialScreen Pty Ltd of Prahran, Victoria, Australia, with a US entity (TrialScreen Inc.) and offices in Boston and London. The business was formed in 2024 when an entity controlled by Dr. Hugo Stephenson, a former Quintiles executive and DrugDev founder, acquired the Opin patient recruitment business from ASX-listed Opyl Ltd for a nominal AU$1, with Opyl retaining a 20 percent equity stake. Patients and physicians can search 100,000+ actively recruiting trials indexed continuously from international registries, check eligibility through an automated screener, and connect with study teams, all without creating an account. TrialScreen is corporately independent of Outcomes4Me. The two were previously queued as one combined HugoScore seed entry but are unrelated companies that happen to operate in adjacent trial-matching space.

The CAIHL read is a two-sided platform. The patient-facing search is unusually low-friction and low-demand: free, global, no registration, no data required to browse. The company's actual customers are sponsors, CROs, SMOs, sites, and academic researchers buying recruitment software and services, including digital triage with AI-assisted eligibility criteria, multi-channel lead generation, and ResearchFriends community building. The patient funnel is the recruitment product. That does not make the search untrustworthy, but it means featured placement, managed-trial prioritization, and conversion incentives deserve the same scrutiny CAIHL applies to any recruitment-funded patient tool.

Evidence Reviewed

CAIHL Profile

  • Who does this AI serve? Hybrid. The search engine is free and openly patient-usable, but the operator is a recruitment business whose paying clients are sponsors, CROs, SMOs, sites, and academics, and the former parent Opyl retains a 20 percent stake.
  • Can patients tell AI is involved? Partial. The privacy policy states the eligibility checker is an automated process whose outcome depends on the information provided and the trial criteria. Researcher pages disclose AI-assisted criteria for registry-sourced trials and deterministic matching for managed trials. How clearly the patient flow labels automation and paid or managed trial placement needs verification.
  • Can patients meaningfully choose? Yes, with funnel caveats. No account or registration is needed to search. Eligibility checks are self-initiated. Contact details go to study teams only when the user chooses to connect, and removal can be requested, though after connection removal requests are forwarded to the study team rather than handled by TrialScreen directly.
  • Can patients correct or challenge what the AI produces? Partial. Access and correction rights exist for personal information, but there is no public workflow to contest an automated eligibility outcome.
  • Does it help patients understand or act? Yes. Global search, condition hubs, eligibility self-screening, direct researcher connection, and research-literacy FAQ content help patients find research options their clinicians may never raise.

Agency Interpretation

The clearest agency value is access without extraction. Most trial-matching tools require an account and a health profile before showing anything. TrialScreen lets a patient or family member search the global registry landscape anonymously and decide for themselves when to disclose anything. For a population told that trials are not an option, or never told about them at all, a free open search engine is a real expansion of the option space.

The unresolved tension is whose trials surface first. TrialScreen earns revenue when sponsors pay for recruitment, runs managed trials with custom-built eligibility criteria alongside AI-annotated registry trials, and features specific research opportunities on its homepage. The privacy policy is substantively reasonable, including no sale of personal information and consent-based sharing, but the published page contains leftover placeholder text and duplicated old and new policy sections, which is a governance hygiene concern for a company handling health-related disclosures. CAIHL treats that kind of sloppiness as a signal worth recording, not as proof of bad practice.

Key Unknowns

  • Whether sponsored, managed, or featured trials are labeled as such in patient-facing search results and ranked differently from registry-only trials.
  • How accurate the automated eligibility checker is, and whether anyone independent has evaluated false negatives that could wrongly discourage patients.
  • What AI models or vendors power the AI-assisted criteria annotation, and whether user data feeds model training.
  • Patient-side language support and accessibility of the search and screening flow.
  • How long patient-provided screening data is retained and whether the 7-year client-file retention language applies to patient records.
  • The full corporate structure across TrialScreen Pty Ltd, TrialScreen Inc., and Opyl's residual 20 percent stake.
  • Whether any ethics or regulatory oversight applies to the recruitment advertising and influencer campaigns described on the researcher pages.

Publication Recommendation

Ready for human review as a draft profile at medium confidence. Identity, ownership, and business model are well verified from official pages and credible divestiture coverage. Confidence would rise with a hands-on walkthrough of the search and eligibility flow documenting AI and sponsorship labeling, clarification of patient data retention, and any independent assessment of matching accuracy. The cleaned-up combined Outcomes4Me/TrialScreen seed entry should be retired in favor of this entry and the separate Outcomes4Me profile.