A haematologist in Japan finishes a clinic and, before the next patient arrives, types a question about treatment sequencing for a case that sits outside the trial evidence. The answer comes from an AI assistant embedded in a community she already uses, alongside colleagues who have seen similar disease. The query is recorded. Multiplied across tens of thousands of physicians, that record is the commercial logic behind the partnership between Singapore's Docquity and Japan's exMedio.

The arrangement places Docquity's clinical assistant, Ask Dx, inside exMedio's existing networks. One is a consultation platform used by more than 80,000 physicians. The other reaches over 60 percent of Japan's haematologists. Oncology comes first, with plans to extend from blood cancers to solid tumours. A wider "One Asia" initiative aims to link more than 500,000 verified doctors across Southeast Asia, Japan and the Gulf. Docquity also describes aggregating real-time physician search data to show educators and developers what clinicians need to know.

The relevance to the NHS is structural and cautionary. Neither company has announced operations in Britain, and nothing here changes a commissioning decision. The interest lies in what the model exposes about a gap in English health policy. Clinicians in the NHS answer difficult questions through a patchwork of national guidance, local formularies, trust pathways, specialist networks and informal messaging groups. That arrangement works because experienced people make it work. It is also invisible to the system, which has no reliable picture of what its clinicians are unsure about.

The 10 Year Health Plan commits the service to a shift from analogue to digital, and ambient voice technology and AI feature prominently. Most of that investment is aimed at documentation and administrative load, which is a sensible place to begin given the workforce strain. Decision support that sits inside the clinical workflow and draws on peer experience has received less attention. In cancer, where waiting times remain under pressure and access to treatment varies between regions, a trusted route to specialist opinion would carry obvious value. Multidisciplinary teams exist, but their capacity is stretched.

The data question deserves more scrutiny than it currently gets. NHS patient data is governed by strict rules, including the national data opt-out and the move towards secure data environments. Clinician query data falls into a gap. It describes the professional rather than the patient, so it escapes much of that protection. Yet it reveals where guidance is ambiguous, where training has lapsed, and which new medicines clinicians are asking about before a commissioning decision has been made. For a life sciences company, that is a demand signal of real worth. For NICE or an integrated care board, it would be useful feedback. If commercial platforms hold it, the NHS ends up buying back insight into the uncertainty of its own workforce.

Regulation adds a further complication. A tool that answers clinical queries may qualify as medical device software, which brings MHRA oversight and the clinical safety standards that NHS deployments must meet. The cross-border promise sharpens the problem. Docquity says its network will account for local differences in practice, and that is the hardest part to deliver. Licensing, formularies and referral norms in Osaka, Riyadh and Manchester differ in ways that matter at the bedside. A British clinician receiving an answer calibrated to another jurisdiction is a patient safety risk, so provenance and localisation need to be visible in the interface rather than buried in terms of service.

Several practical points follow for those making decisions. NHS leaders could usefully find out where their clinicians already turn for answers and whether those sources are sound. Policymakers should clarify the status of data generated by clinician behaviour, ideally before a market forms around it. Health-tech firms should note that the exMedio deal depends on reach that took years to build. Distribution through trusted professional communities appears to count for more than the quality of the algorithm.

The NHS may never host a platform like Docquity. Whether it treats the questions its clinicians ask as public infrastructure is a decision it will make by default if it does not make it deliberately. Japan and Southeast Asia are showing what happens when someone else builds it first.

‍