Validation of adapted daily dose definitions for hospital antibacterial drug use evaluation: a multicentre study

The WHO/ATC (Anatomical Therapeutic Chemical) index DDD (WHO-DDD) is commonly used for drug consumption measurement. Discrepancies between WHO-DDD and actual prescribed daily doses (PDD) in hospitals have prompted alternative dose definitions adapted to doses recommended in hospital practice guidelines [recommended daily doses (RDD)]. In order to validate RDD we performed modified point prevalence surveys in 24 acute care hospitals and recorded 20620 PDD of antibiotics given to 4226 adult patients on the day of the survey and the 6 preceding days. We calculated RDD and WHO-DDD and compared them with PDD. The rate of RDD corresponding to PDD was higher than the corresponding rate for WHO-DDD (pooled data, 55% versus 30%) and the differences were similar across the hospital sample, but varied according to drug/drug class, route of administration, indication and renal function. RDD underestimated actual consumption by 14% overall, while WHO-DDD overestimated total antibacterial consumption by 28% (pooled data; median values RDD -10% versus WHO-DDD +32%). The deviations of estimated from actual drug use volumes were largest for β-lactams (RDD -11% versus WHO-DDD +49%), in particular for penicillins (-11% versus +64%), if WHO-DDD were used. Hospital antibiotic consumption surveillance systems using current WHO-DDD should address the uneven discrepancies between actual prescribing and consumption estimates according to drug class that may lead to misclassification in benchmark analyses. We recommend using validated RDD as a supplementary measure to the WHO-DDD for detailed analyses.