Repeated nationwide point-prevalence surveys of antimicrobial use in Swedish hospitals: data for actions 2003-2010.

This study sought to analyse antimicrobial pressure, indications for treatment, and compliance with treatment recommendations and to identify possible problem areas where inappropriate use could be improved through interventions by the network of the local Swedish Strategic Programme Against Antibiotic Resistance (Strama) groups. Five point-prevalence surveys were performed in between 49 and 72 participating hospitals from 2003 to 2010. Treatments were recorded for 19 predefined diagnosis groups and whether they were for community-acquired infection, hospital-acquired infection, or prophylaxis. Approximately one-third of inpatients were treated with antimicrobials. Compliance with guidelines for treatment of community-acquired pneumonia with narrow-spectrum penicillin was 17.0% during baseline 2003-2004, and significantly improved to 24.2% in 2010. Corresponding figures for quinolone use in uncomplicated cystitis in women were 28.5% in 2003-2004, and significantly improved, decreasing to 15.3% in 2010. The length of surgical prophylaxis improved significantly when data for a single dose and 1 day were combined, from 56.3% in 2003-2004 to 66.6% in 2010. Improved compliance was possibly the effect of active local feedback, repeated surveys, and increasing awareness of antimicrobial resistance. Strama groups are important for successful local implementation of antimicrobial stewardship programs in Sweden.

Timeliness of case reporting in the Swedish statutory surveillance of communicable diseases 1998--2002.

To prepare for a new communicable disease act in Sweden, we performed an in-depth analysis of the performance of the present notification system (1998--2002). Four diseases were selected for analysis (meningococcal infection, salmonellosis, infection with penicillin-resistant pneumococci and tularaemia). Each step in the double notification flow (clinical and laboratory notifications) was studied and paper-based and electronic notifications compared. More than 15 possible single dates in the notification flows were analysed for 27,000 cases. The shortest notification flows (from date of sample collection to when the first notification was received at SMI--clinical or laboratory notification) were seen for meningococcal infections and salmonellosis (median 7 d) and the longest for tularaemia (11 d). Laboratory notifications were faster than clinical notifications, and electronic reporting 1.5-1.7 times faster than paper notifications. Further gains should be expected from the new electronic reporting system (SmiNet2), which will allow also clinical reporting electronically over the internet.