Who will make personalised medicine work?
The CPM was delighted to host Maxine Mackintosh as one of its first affiliate members. In this blog Maxine makes some very pertinent insights about the conditions needed to make personalised medicine work. Maxine has now moved across the pond but we do hope she will return to work with us at the CPM at some point in the future!
Personalised medicine is often framed as a scientific frontier, but it is also an evidential, organisational, and ethical one. The question is no longer only how to generate better science, but how to recognise, test, and use new forms of evidence in real systems, and to do so in ways that are trustworthy and equitable. As medically relevant evidence increasingly emerges beyond traditional academic settings, the people able to move between academia, implementation, and public legitimacy become evermore important. They are often the ones helping to make rigour operational, turn equity into design choices, and build trust in how innovation is introduced and governed.
I am probably predisposed to notice this, having spent much of my career so far trying, not always neatly, to move across different roles, disciplines, and perspectives. Trying to bridge those worlds has made me more aware of how much sits between them, and how difficult it can be to work well in that space. But the broader value of cross-sector, interdisciplinary, boundary-spanning work has been increasingly on my mind in recent years: through a Sciana Fellowship and its emphasis on systems leadership for health, recently reading a very validating book called Range which celebrates generalists in a world of specialists, and through a BMJ Leader piece by a friend and colleague making the case for managerial academic career (in essence, for people who can speak implementation and operations as fluently as they speak academia and evidence). That feels especially timely in personalised medicine, where the task is not to loosen standards, but to make sure rigour moves with evidence into the places where medicine is now learning, and that there are people able to steward that well.
Why this matters now
Part of the reason this matters more urgently now is that the environment of evidence itself is changing. Evidence-based medicine was built for a world in which evidence was relatively scarce, expensive to generate, slow to accumulate, and largely produced in a limited number of formal settings. Academic centres ran studies, journals published them, guideline bodies synthesised them, and clinicians then applied the results to patients. That architecture brought enormous gains, and still does. But it is no longer the only world medicine inhabits.
Patients can now generate continuous physiological and behavioural data through wearables, sensors, apps, home testing, and patient-reported measures. Clinicians work within electronic records that contain large volumes of routine outcome data. Digital platforms can vary interventions at scale and learn from the results. Information from daily life, behaviour, environment, service use, and treatment response can increasingly be linked together. AI systems can interrogate those data at a speed and scale that no individual clinician or researcher could manage unaided.
That does not mean every new data point is meaningful, nor that every new data source deserves the status of evidence; assuming so has caused all manner of problems. But it does mean that clinically relevant insights are increasingly emerging outside the traditional academic centre. That matters not only because it changes where evidence comes from, but because it raises harder questions about who gets counted, who is left out, and who is equipped to judge what should shape practice.
Why the old division of labour is becoming harder to defend
What is becoming harder to ignore is that some of the people best placed to improve healthcare are still not the ones most clearly supported to produce knowledge about it. Those working close to delivery, or using services, often see first how systems behave under pressure: where pathways fail, where evidence stalls, where incentives distort decisions, where trust is won or lost, and what it actually takes to make innovation usable. Yet they are still too often cast as implementers of knowledge produced elsewhere, rather than as contributors to the generation, interpretation, and translation of knowledge itself.
The reverse problem exists too, as academic work can remain oddly insulated from the operational conditions in which it will eventually have to function. In personalised medicine, that separation becomes harder and harder to defend. The old distinction between those who generate knowledge and those who merely apply it looks increasingly unconvincing. If the ultimate goal is to improve health for patients, service users, and citizens, then healthcare has to build far stronger routes for people in real-world roles to become more academically grounded, while also expecting academic work to remain more answerable to practice.
Personalised medicine makes the problem hard to ignore
Take genomics, often held up as one of the great scientific promises of personalised medicine. It is certainly that, but it is also a useful reminder that discovery alone is rarely enough. Even one of the most scientifically advanced areas of personalised medicine succeeds or fails not only through breakthroughs in the lab, but through pathway design, workforce readiness, interoperability, commissioning, and trust. In other words, discovery is only part of the story.
The NHS Genomic Medicine Service offers a good example in that the scale of testing and the expansion of the National Genomic Test Directory are significant achievements. But the less glamorous work matters just as much, which lies in standardisation, data flowing across systems, and innovation infrastructure able to support adoption in everyday care. A review of integrating genomic medicine into mainstream NHS care made similar points. The barriers to real clinical utility are often as practical and systemic as they are scientific: referral-dependent pathways, unclear eligibility criteria, gaps in workforce capability, cost pressures, unequal access, and ongoing questions around data management and privacy.
Those are not simply operational obstacles, they are also issues of fairness, confidence, and public legitimacy. Anyone who has worked in or alongside delivery will recognise some of that; I certainly do from my own time at Genomics England, where the scientific promise was always inseparable from the human, technical, and organisational capability around it.
The people we need go in both directions
We need more people in operational, managerial, and service-facing roles to be supported to become more academically grounded: more confident in evidence, more fluent in method, and better able to distinguish a promising signal from a justified change in practice. As that BMJ Leader piece argued, many managers already bring research literacy, operational knowledge, and system-level insight, but lack the formal structures, incentives, and progression routes that are much more clearly established for clinical academics. The point is not simply that managers should be “allowed in” to academic space, but that healthcare underuses the people who often understand the system best.
The movement has to go the other way too, of course. We also need more academics in personalised medicine whose work is shaped earlier, and more seriously, by the conditions of implementation. A method, biomarker, genomic insight, or data tool may be scientifically compelling and still fail in practice if it does not reckon with workflow, time, incentives, governance, interoperability, or the distribution of expertise across services. It is one thing to understand that in theory, and quite another to feel the pressure of delivery itself, which is always, among other things, humbling.
These boundary-spanning people are also often the first to spot the early signals of inequity. Because they work across worlds, they are frequently the ones who notice when certain groups are missing, when design choices may be reinforcing unfairness, when trust is being assumed rather than earned, and when technically impressive work is at risk of losing social legitimacy. That reciprocity is what I find most compelling: the possibility of building more people, roles, and habits that allow evidence and implementation to shape one another earlier and more honestly.
So what should we take from this?
The most practical implication is that personalised medicine needs to get much better at recognising and backing the work that sits between established categories. Too often, the people trying to connect evidence to delivery, operational reality to academic method, or technical promise to public interest are left to do that work by instinct, goodwill, or sheer persistence. If we think they matter, then they need more than recognition. They need clearer routes, stronger backing, and institutions willing to treat that work as serious.
I feel very lucky to have often existed in environments where this kind of work was possible. But it has become clearer to me over time that this has not been the experience of many others. That matters, because if this work depends too heavily on energetic individuals or unusually supportive settings, then it remains fragile.
A second implication is that the goal should not simply be better collaboration between separate groups, but more shared literacy across them. People in real-world roles should be able to engage more confidently with evidence, method, and uncertainty. Academics, in turn, should be expected to stay closer to the realities of implementation, service design, and public trust. The point is not to collapse every boundary, but to make those boundaries less limiting than they often are now.
These people also tend to act as early course-correctors. They can see where a pathway looks elegant on paper but awkward in practice, where a promising innovation is subtly excluding some groups, or where confidence is being assumed before it has really been earned. That kind of perspective is what allows personalised medicine to mature well.
And perhaps that is the real test. Not simply whether personalised medicine is producing more science, more tools, or more technical sophistication, but whether it is building the people, habits, and institutions that allow those advances to be used wisely, fairly, and with enough humility about what real-world care actually demands.
What the Centre for Personalised Medicine can contribute
Perhaps that is why spaces like the Centre for Personalised Medicine matter so much. Not because they have all the answers, or because every difference in perspective can be neatly resolved, but because they can act as a kind of third space: somewhere researchers, practitioners, patients, and publics can come together outside the usual constraints of role, discipline, or institutional hierarchy. In spaces like that, people can sometimes speak a little more openly, ask better questions, and encounter one another first as humans rather than as job titles or forms of expertise.
For me, that is one of the most valuable things a centre like CPM can offer. Not simply a platform for excellent research, but a place where ideas can be tested against practice, where practice can be more firmly grounded in evidence, and where relationships of trust and curiosity can develop across boundaries that too often stay fixed.
If personalised medicine is to mature well, it will be judged not only by what it discovers, but by whether it can build the people, cultures, and institutions able to carry those discoveries into care fairly. It will need more spaces like CPM: places that do not flatten difference, but make it easier to work across it with honesty, humility, and purpose.
Maxine Mackintosh






















