This commentary makes a narrow, well-evidenced argument: dosing recommendations built from healthy volunteers and selected pre-registration populations may not transfer to physiologically distinct groups, and low- and middle-income countries contain most of the groups nobody studied.
Two South African examples carry it. A metformin-dolutegravir interaction study shows body composition altering an assumption built elsewhere. A colistin pharmacokinetic study shows disease severity doing the same. In both, the published dose was not wrong so much as derived in people who differ from those receiving it. The authors' proposal is procedural: drug development and post-registration programmes should carry an explicit translation plan, using data from the target population to confirm, refine or revise the dose.
The argument lands directly in Indian practice, where the gap is not exotic. Lower average body weight and different body composition at a given weight, a high prevalence of undernutrition, tuberculosis co-treatment with rifampicin's induction running through everything, and widespread renal impairment that was an exclusion criterion in the registration trial. The usable habit is to ask, for any drug you prescribe at a standard dose in an atypical patient, where that number came from and whether anyone like this patient was in it. Therapeutic drug monitoring, where it exists, is how that question gets answered rather than merely asked.
- Ask which population a standard dose was derived in before applying it to an atypical patient.
- Body composition, not just weight, changes exposure - particularly for lipophilic and hydrophilic extremes.
- Disease severity alters clearance in ways healthy-volunteer studies cannot capture.
- Use therapeutic drug monitoring where available rather than assuming the label dose transfers.
- Rifampicin co-treatment silently invalidates many standard doses in Indian practice.
Why it matters
Most of the world is prescribed doses derived from people it does not resemble, and nothing in the label flags that.
Don't overread it
A commentary illustrating a principle with two examples, not new dosing data for any drug.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for clinical pharmacology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free