Comparison of different antibiotic consumption measurement methods in large multidisciplinary hospital

Abstract number: p1476

Pujate  E., Apine  I., Dumpis  U.

Objectives: 

Antibiotic selection pressure is determined by the total amount of antibiotics, number and density of patients treated with antibiotics in the particular geographical area. Several antibiotic consumption detection methods should be combined in the hospital setting. Our objective was to evaluate efficacy of different approaches in large multidisciplinary hospital.

Methods: 

Point prevalence studies were repeated annually at 2002–2004 in Stradins University Hospital (1000 beds) in Latvia. All patients receiving antibiotics on the day of the survey were identified and their medical records were reviewed. Data on antibiotics, dose and route of administration were collected. In addition, annual data on antibiotics dispensed to the departments were collected from pharmacy. Total used grams for each antibiotic were expressed into defined daily doses (DDD-WHO). Bed days (BD) and admission days (AD) were used as denominators.

Results: 

Table 1 Total use of antibiotics in Stradins University Hospital 2002–2004.

The most commonly used antibiotic groups in the pharmacy study were 1st generation cephalosporins (13.35 DDD/100 BD in 2002, 11.6 in 2003, 10.8 in 2004) and penicillin's with extended spectrum (11.20, 11.98, 13.15) followed by fluoroquinolones (6.26, 8.13, 8.69) and metronidazole (4.42, 4.59, 5.61). There was no significant difference between distribution of different antibiotics from prevalence and pharmacy studies if calculated in DDDs. In contrast, distribution of antibiotics calculated per patient in the prevalence study was quite different; 1st generation cephalosporins (8.71%, 8.63%, 5.84% in 2002, 2003, 2004 respectively) and fluoroquinolones (3.05%, 6.69%, 5.63%) with smaller proportion of extended spectrum penicillins (4.36%, 3.88%, 3.93%) and metronidazole (4.03%, 3.56%, 4.68%).

Conclusions: 

There were no differences in the distribution of antibiotics calculated in DDDs per bed days and admissions. Distribution of antibiotics in annual pharmacy studies and point prevalence studies if calculated in DDDs were also similar. In contrast, the prevalence data expressed as a proportion of patients with selected antibiotics showed quite different distribution. Studies using only DDDs may overestimate use of certain antibiotic groups in our setting where WHO DDDs are significantly different from actual PDDs used.