Separating signal from narrative in healthcare innovation.
Diligence Dx is the applied layer of the HYPE Index, a stage-aware, gap-centered methodology for scoring healthcare AI claims across companies, products, consumer influencers, and research papers.
Healthcare innovation moves faster than evidence. Funding narratives, media velocity, and category enthusiasm often run ahead of clinical validation, regulatory clearance, and real-world performance. The signal is the gap, not the evidence. Diligence Dx applies a structured framework to ask a different question of every claim: what is verified, what is still narrative, and where is the gap?
Built for physicians, investors, journalists, researchers, and the patients and consumers on the receiving end of healthcare AI claims.
Consider how a single biological insight travels through the public information environment. A laboratory publishes a calibrated paper showing that a biomarker is associated with a disease outcome in a specific cohort. A company writes a marketing page that cites the paper, dropping the qualifier. A founder, in a podcast interview, says the biomarker predicts the disease. A clip is shared. The viral caption reads: scientists discovered the marker that determines whether you will get this disease.
The signal does not survive the chain.
By the time the claim reaches a consumer, the cumulative gap is large. The HYPE Index scores each link on the same rubric, which is what makes the propagation of miscalibration measurable.
- Stage-aware: a frontier laboratory and a commercial company are held to different evidentiary expectations.
- Four entity types on a single rubric: companies, products, consumer-facing influencers, research papers.
- Sixteen criteria across four axes, each scored on a five-point anchored scale.
- Two numbers reported: the E score (the headline gap) and the four-axis average (the diagnostic context).
- Calibration Gap threshold calibrated against retrospective backtest, not stipulated.
The methodology has been hardened by a twelve-persona stress-test panel. The panel includes a working clinician asking whether the rubric can be applied between admissions, a psychometrician asking whether anchored descriptors hold their measurement-theoretic shape, and an investigative journalist asking whether scored outputs survive in published narrative. It also includes a senior peer reviewer testing whether the research-paper scoring guide captures the gestalt judgment experienced reviewers actually use, a patient advocate asking whether the rubric makes visible the calibration failures that matter to patients on the receiving end of healthcare AI claims, and a regulatory affairs director checking whether the rubric uses language consistent with how regulators evaluate the same calibration questions. Six other personas round out the panel. Each holds a distinct diagnostic question. Full panel synthesis is documented in the v1.3 methodology paper.
Diligence Dx is the editorial layer. The HYPE Index is the scoring layer. Methodology, scoring rubric, and editorial governance are published in full so readers can verify the framework against their own judgment.
The HYPE Index is not investment advice. It is not a regulatory instrument. It does not authorize, approve, or clear any entity. The methodology paper defines how to score; the application papers define what was scored.
The v1.3 methodology paper is available to qualified reviewers on request.
Send a note below to request the v1.3 methodology paper, ask a question, or help shape the instrument.
The HYPE Index is editorial commentary on public information, in the tradition of Consumer Reports, Morningstar, and the major credit rating agencies.
Diligence Dx scoring is editorially independent of any investment activity by affiliated entities.
Diligence Dx publishes opinion-based evaluations using a transparent methodology applied to disclosed facts. Evaluations are not investment advice and do not constitute recommendations to buy, sell, or hold any security.