A 2026 prospective pilot study, published in Cureus, evaluated physician-perceived value of DR.INFO v1.0 in routine clinical practice. The results were positive — but interpreting them correctly requires understanding what the study measured, what it did not measure, and what stronger evidence would look like.
Why the DR.INFO Study Matters
Most clinical AI tools launch with marketing claims and user testimonials. DR.INFO has published a peer-reviewed pilot study with a described methodology — that places it ahead of many competitors in the evidence hierarchy, even if the study has significant limitations.
The study matters because it represents one of the first attempts to formally evaluate a European clinical AI reference tool in routine clinical practice, with a structured protocol and reported outcomes.
What the Study Design Was
The study was a prospective, single-arm pilot conducted across Portuguese healthcare institutions. Twenty-nine physicians and medical students were recruited to use DR.INFO v1.0 over five working days within a two-week period. Participants used the tool during routine clinical practice — not in a simulated or controlled laboratory environment.
Sixteen of the 29 participants completed the final evaluation. The primary outcomes were self-reported measures: perceived time saving, perceived decision-making support, and Net Promoter Score (NPS). Data was collected through structured questionnaires.
What the Results Showed
Among the 16 participants who completed final evaluation, the reported outcomes were positive.
Perceived time saving: 4.27 out of 5. Perceived decision-making support: 4.16 out of 5. Net Promoter Score: 81.2 — indicating that participants would strongly recommend the tool.
These scores suggest that physicians who used DR.INFO found it subjectively useful for saving time and supporting their clinical work. The NPS of 81.2 is exceptionally high — above the threshold typically associated with strong product-market fit.
What the Results Do Not Prove
The study measured perceived value — not objective clinical outcomes. This is an important distinction.
No diagnostic accuracy measurement. The study did not evaluate whether DR.INFO's answers were factually correct or clinically appropriate. Perceived decision support ("I felt this tool helped me make decisions") is not the same as measured decision quality ("this tool's recommendations were accurate").
No patient outcome measurement. The study did not track whether using DR.INFO led to better patient outcomes, fewer diagnostic errors, or improved prescribing quality.
Small sample. Twenty-nine participants recruited, 16 completers. This is too small to draw generalisable conclusions about the broader physician population.
Single-arm design. No comparator group. Without a control group using alternative tools (or no tool), we cannot attribute the perceived benefits specifically to DR.INFO versus the general effect of having any additional clinical reference available.
Short duration. Five working days. Novelty effects can inflate perceived usefulness in short-duration studies — clinicians may rate a new tool highly during initial exploration, with satisfaction declining as the novelty fades.
Self-reported outcomes. All primary measures were subjective self-reports. Self-reported time saving may not correspond to actual time saved when measured objectively.
The study authors are transparent about these limitations, concluding that "larger controlled studies with objective outcome measures and independent accuracy verification are needed." This intellectual honesty is commendable and should be noted.
Why Perceived Usefulness Still Matters
Despite its limitations, perceived usefulness is not meaningless. If physicians find a tool helpful in daily practice, they will use it. Adoption — regardless of rigorous outcome data — drives real-world clinical AI integration. The NPS of 81.2 suggests that DR.INFO v1.0 generated genuine enthusiasm among its pilot users.
The parallel from other fields: many technologies achieve widespread adoption based on perceived utility long before randomised controlled trials confirm objective benefit. The question is whether perceived utility translates into sustained use and measurable clinical value.
What Stronger Evidence Would Look Like
A stronger evaluation of DR.INFO — or any clinical AI search tool — would include: larger sample sizes (hundreds of participants across multiple institutions), controlled design (randomised comparison against alternative tools or standard practice), objective outcome measures (diagnostic accuracy, time-to-information measured by system logs, prescribing quality metrics), longer duration (months, not days), and independent evaluation (conducted by researchers without commercial relationship to the product).
The HRA-listed Medwise pilot study — comparing AI guideline search against manual hospital intranet search — represents one model for this kind of controlled evaluation in the UK context.
What This Means for Clinical AI Adoption
For clinicians evaluating DR.INFO or any clinical AI tool: early pilot data showing positive perceived usefulness is encouraging but not definitive. The tool may genuinely help — the pilot suggests clinicians think it does. But "clinicians perceive it as useful" and "the tool objectively improves clinical decisions" are different claims requiring different evidence.
For clinicians, the practical approach: try the tool, verify its outputs against primary sources, assess whether it fits your clinical workflow, and make your own judgment about its usefulness in your practice.
For clinicians looking for a UK-focused clinical knowledge platform with cited answers, calculators, and exam preparation: iatroX is built around the UK clinician workflow. UKCA-marked, MHRA-registered. Free.