For most of medical history, treatment has been statistical: this drug works for most people with this diagnosis, at this dose, and if you are in the unlucky minority it fails for, the doctor tries the next one. Personalised medicine proposes to invert the process — to read the patient’s own biology first, from tumour genetics to drug metabolism, and choose the treatment that the individual’s body is most likely to answer. After decades as a conference slogan, the approach is now producing everyday clinical wins.
Where it already works
Oncology led the way. Many cancers are now tested at diagnosis for specific mutations, and the results steer therapy: drugs that block particular growth signals can produce dramatic responses in patients whose tumours carry the matching alteration, while sparing others a treatment that could never have helped them. Pharmacogenomics — testing how a patient’s genes affect drug processing — now informs prescribing for certain blood thinners, antidepressants and pain medicines, where the difference between a slow and fast metaboliser is the difference between a dose that fails and one that harms.
Reporting in journals covered by Nature’s personalised medicine coverage tracks the frontier moving further: cancer vaccines manufactured for a single patient’s tumour, gene therapies for rare diseases designed around one family’s mutation, and polygenic risk scores that flag who should be screened years earlier than population guidelines suggest.
The obstacles are mundane, not scientific
On oneonenews.com our health team hears the same bottlenecks from clinicians: testing costs and turnaround times, data systems that cannot share results between hospitals, reimbursement rules written for the old model, and a workforce that needs new skills to interpret genomic reports. Equity looms largest — personalised medicine built on datasets drawn mostly from a few populations risks working best for exactly those populations.
Our Health analysis also notes the privacy frontier: a genome is the most identifying document a person owns, and the rules governing who may see it — insurers, employers, researchers — lag the technology that generates it.
The clinic of the near future
The destination is not bespoke medicine for every ailment, but a steady expansion of the questions doctors ask before prescribing: not just ‘what disease?’ but ‘whose disease, with which markers, in which body?’ The stethoscope took decades to become universal. The genomic report is on the same journey — slower than the hype, faster than the sceptics expected, and already changing outcomes for patients whose treatment finally fits their biology.
What patients should ask
Clinicians suggest patients become participants in the transition with three questions. Is there a test that would change my treatment — and has it been offered? If a drug is prescribed in a class known for genetic variation in response, does my record carry the relevant result, or should it? And where will my data live, who can see it, and can I take it with me between providers? None of these questions requires a science degree, and all three nudge the system toward the medicine it claims to be building. Personalised medicine’s final ingredient, as with every era of medicine before it, is an informed patient asking the newly possible question at the right moment.
