DATA: bias study

The file is in a .cvs format outputted through STATA version 16.0.

The dataset is organised in the long format with the combination of the variables id(), species(), and antibiotic() identifying a unique record.

A codebook is uploaded as a separate file indicating variable labels and value labels.

The project assessed the difference in point prevalence of antimicrobial resistance (AMR) when using routine laboratory-based surveillance and results from an active population-based survey (population-based surveillance). The data are restricted to the prevalence of AMR in isolates of Escherichia coli and Klebsiella pneumonia cultured from urine samples. For the laboratory-based surveillance, all available samples were included, while the population-based surveillance included the urine from all patients presenting with sign and/or symptoms of a urinary tract infection in either the out-patient or inpatient setting. Laboratory-based surveillance used the Vitek methodology to assess resistance, while the population-based surveillance used the conventional culture methods using MacConkey agar plates incubated for 18 hours. Resistance was defined by applying the CLSI breakpoint defined in 2012. The data from both sources covered exactly the same period (1 April 2014 to 31 May 2015) in which there were no reports of outbreaks in the hospital setting.