EVIDENCE LIBRARY
Taiwan Medication Audit Produced Different Inappropriate-Prescribing Prevalence Under Three Criteria Sets
A large claims analysis found that screening definitions changed the measured prevalence of potentially inappropriate medication use. A flagged benzodiazepine prescription was not a documented injury.
A large claims analysis found that screening definitions changed the measured prevalence of potentially inappropriate medication use. A flagged benzodiazepine prescription was not a documented injury.
Three definitions applied to one population
The study examined ambulatory records for more than 1.16 million older Taiwanese adults in 2009. It applied the 2012 Beers, PIM-Taiwan, and PRISCUS criteria separately.
The proportion with at least one potentially inappropriate medication ranged from 66.9% to 86.2%, depending on the criteria. Benzodiazepines accounted for many leading entries under the Taiwanese and PRISCUS approaches, while another drug class predominated under Beers.
Screening is not adjudication
Different lists have different inclusions, exclusions, and dose or context rules. Their prevalence estimates therefore answer different operational questions. Claims also lack important clinical details such as functional status, living situation, and some reasons for prescribing.
The analysis did not determine that every flagged prescription caused an adverse event or that the criteria were equally predictive of harm.
For an individual assessment
Identify the criteria version and exact rule before calling a prescription potentially inappropriate. These historical criteria are not automatically a current legal standard. The study demonstrates sensitivity to definition, while an individual clinical opinion requires indication, dose, duration, alternatives, and the observed course.
Source paper
Original analysis of the supplied journal article, identified as PMC4636615.