Amount of alcohol-based hand rub used and incidence of hospital-acquired bloodstream infection in a danish hospital

acquired if it began more than 48 hours after hospitalization and was not present at or incubating before hospital admission; BSIs in patients readmitted within 7 days after microbiologically confirmed HCR-BSI were not considered new cases of HCR-BSI. BSIs were identified from the microbiological laboratory information system and included primary and secondary cases. The Danish civil registration numbers of patients with microbiologically confirmed HCR-BSI were used to review electronic patient records for signs and symptoms of BSI. Data on the number of bed-days, the times of admission and readmission, and the number of admissions were compiled from the hospital’s administrative database. The amount of hand rub used per 1,000 bed-days (represented on a log scale) was analyzed jointly with the number of new cases of HCR-BSI per 1,000 bed-days. We used a vector autoregression model, with the amount of hand rub used and the number of incident HCR-BSIs in the preceding 2 months