A pilot study of risk adjustment for benchmarking antibiotic use between hospitals in Sweden

There is no established standard for comparing overall antibiotic use between hospitals taking patient characteristics into account. The objective of this study was to investigate whether there is a correlation between surrogate markers for patient morbidity, namely case mix index (CMI), mean length of hospital stay (LoS) and mean cost per admission, and antibiotic use in a sample of Swedish hospitals. All primary and secondary hospitals in three counties with high and three counties with low consumption of antibiotics were selected. Data from 16 hospitals were included. A regression analysis was used to evaluate whether there was a linear trend between defined daily doses (DDD) of antibiotics per 100 bed-days and the surrogate markers for morbidity. No correlation could be found between any of the measures of morbidity and total antibiotic consumption. However, a correlation was found between CMI and the proportion of narrow-spectrum antibiotics: the higher the CMI, the lower the proportional use of β-lactamase-sensitive penicillins. In conclusion, it was found that CMI, mean LoS and mean cost per admission did not appears to be useful factors to adjust for when comparing antibiotic use in this subset of primary and secondary care hospitals. Based on this limited study, we suggest that DDD/100 bed-days can still be used as an appropriate metric to benchmark antibiotic use in primary and secondary hospitals until a better marker for variation of patients and activities is identified.