WHO defined daily doses versus hospital-adjusted defined daily doses: impact on results of antibiotic use surveillance.

OBJECTIVES To investigate effects on surveillance results of hospital antibiotic use when WHO defined daily doses (WHO DDDs) are adjusted to doses recommended for hospitalized patients [hospital-adjusted defined daily doses (haDDDs)]. METHODS Data for antibiotic use in 2006-11 for all 22 Norwegian Health Enterprises were analysed with both WHO DDDs and haDDDs as numerators. The haDDDs were determined from recommendations given in regional and national guidelines on antibiotic use in hospitals. The two ways of calculating the amount of antibiotic use were compared, with either the number of bed days (BDs) or the number of discharges as the denominator. The drug utilization 90% methodology was applied for ranking the use of the various antibiotics. RESULTS DDD adjustments altered the figures for total antibiotic use from 67.1 WHO DDDs/100 BDs to 49.3 haDDDs/100 BDs (-26.4%). The most marked difference was found for penicillins: 31.1 WHO DDDs/100 BDs versus 13.4 haDDDs/100 BDs (-56.8%). The corresponding figures for broad-spectrum antibiotics were 17.3 and 15.5 (-10.4%), respectively; for these antibiotics, the conversion changes varied significantly between institutions, from -16.7% to -3.3%. Ranking antibiotic use based on haDDDs resulted in higher positions for metronidazole, cefuroxime, cefotaxime and cefalotin/cefalexin compared with the WHO DDD-based ranking, where the penicillins dominated. CONCLUSIONS The low-set WHO DDDs for penicillins caused skewed surveillance results that concealed the real magnitude of broad-spectrum antibiotic use and distorted interhospital comparisons. For surveillance of antibiotic use in hospitals, WHO DDDs should be supplemented with haDDDs.