Design: Retrospective health record analysis combined with population genomic data
Population: 1,157,359 patients at a Singapore tertiary hospital, 2014 to 2021
Primary outcome: Prevalence of pharmacogenomically implicated prescriptions
Effect: 38.1% to 43.0% per year received one; estimated 18.4% could have prescriptions modified

Researchers combined 2014 to 2021 prescribing records for 1.16 million patients at a major Singapore hospital with whole-genome data from 9,051 Singaporeans. They counted medicines with strong pharmacogenomic guidance (CPIC level A or A/B).

Each year, 38.1% to 43.0% of patients with prescriptions received at least one such medicine. Omeprazole, statins and tramadol were the most common, and CYP2C19, CYP2D6 and SLCO1B1 the genes most often involved. Indian and Malay patients received these medicines at younger ages than Chinese patients. Based on variant frequencies, an estimated 18.4% of patients might have had a prescription changed by pre-emptive testing.

The population includes a substantial Indian community, so the pattern is relevant to Indian prescribing. The estimate is modelled, not observed, and testing is not routinely available in most Indian settings.

Omeprazole, statins and tramadol were the commonest prescribed medicines carrying pharmacogenomic guidance
Consider CYP2C19 and CYP2D6 effects when a patient has unexpected toxicity or no response
Where pharmacogenomic results exist, check them before prescribing affected drugs
Pre-emptive testing is not yet routine in India; its value here is modelled, not proven

Why it matters

It shows pharmacogenomics touches everyday drugs, not just specialist ones.

The statistics, in plain English

The 18.4% figure combines prescribing data with population variant frequencies; it is an estimate of potential change, not a measured improvement in outcomes.